Communications

Electroactivity Enhancement by Redox were produced by palladium-catalyzed Stille coupling

Matching in Cobalt Salen±Based Conducting methods (Scheme 1) followed by condensation with ethyl-
Polymers** enediamine.[6a] Monomeric cobalt complexes 3 and 6 were
prepared by standard methods[6b] as black microcrystals
By Richard P. Kingsborough and Timothy M. Swager* from dimethylformamide (DMF) solutions under argon.
Polymeric films of 3 can be readily oxidatively deposited
The preparation of polymer-modified electrodes that on platinum electrodes from a saturated monomer solution
function as electrocatalysts, chemical sensors, and electro- ([3] ca. 0.1 mM) in 0.1 M Bu4NPF6 /MeCN by scanning be-
chromic displays remains an active field of research.[1] We tween ±0.75 V and 0.85 V (Fig. 1).[7] The cyclic voltammo-
are currently engaged in the design of new conducting gram (CV) of a very thin film of poly(3) reveals that the ra-
polymer architectures containing transition metals in well- tio of metal- to organic-centered redox waves is
defined coordination environments.[2] Our interest in these approximately 1:2 (Fig. 2a). Upon growing a film suffi-
materials is motivated by the wealth of desirable character- ciently thick to connect two interdigitated electrodes, we
istics endowed by transition metals for the development of see that the wave attributed to the Co2+/3+ redox process is
sensory materials that should be responsive to anions, much smaller and a much more broad organic-based elec-
Lewis bases, and small molecules (CO, O2, NO, etc.).[3,4] trochemical process centered at 0.45 V is apparent, with a
Polymers based on transition metal±containing N,N¢- ratio of metal- to organic-centered redox waves of approxi-
ethylenebis(salicylidenimine) (salen) complexes have been mately 1:5 (Fig. 2b). The polymers display a broad, poly-
previously prepared by the oxidative polymerization of thiophene-like CV trace, indicative of linear polymeriza-
monomeric M(salen) complexes.[5] In all but a few of these tion through the thiophene substituents, as shown in
polymers,[5d,5j] the electrochemistry of the films displays li- Scheme 2. The polymer structure was indirectly confirmed
gand-based redox processes at high potentials that are con- by scanning a number of related transition metal com-
siderably removed from the expected metal-based electro- plexes of the methyl-terminated salen ligand, N,N¢-ethyl-
activity. These films generally lacked metal-centered enebis(5-(5-methylthienyl)salicylidenimine), to the same
electroactivity; however, thin Co(salen) films did display polymerization potentials. The absence of any polymer
redox processes attributed to the Co2+/3+ couple, which di- deposition (Scheme 3) supports the proposed structure and
minished as the films became progressively thicker.[5d] We similar results have been reported in studies concerning the
have targeted conducting polymers that incorporate the oxidative polymerization of M(salen) complexes.[5]
metal centers as part of the conduction pathway and tightly To account for the decreased electroactivity of the cobalt
couple the conductivity to the redox characteristics of the center in the thick films relative to the thin ones we postu-
metal center. We believe that an organic polymer redox po- late that when the polymer is in the reduced state it effec-
tential that is the same or below that of the metal center tively acts as an insulator, preventing the Co2+/3+ redox pro-
should create materials with enhanced electron mobilities, cess from occurring in the majority of the polymer
and thereby form optimized sensory and catalytic materi- structure. Those centers being oxidized or reduced are
als. In this contribution we demonstrate that tuning the re- most likely to be very close to the electrode surface. We be-
dox potential of the organic polymer backbone to be sim- lieve that the remaining cobalt centers are irreversibly oxi-
ilar to that of the Co2+/3+ redox potential creates enhanced dized to Co3+ during the polymerization process while the
electroactivity of cobalt centers embedded in the polymer polymer is in its oxidized form, as it is more positive that
backbone. This intimate coupling of the metal and polymer the Co2+/3+ redox couple. This has been observed by Hor-
in the conduction process can impart sensitivity to specific witz and Murray.[8]
analytes, and we further demonstrate that Lewis bases such In situ conductivity experiments on 5 mm interdigitated
as pyridine and substituted pyridines affect the polymers' microelectrodes[9] (Fig. 2c) reveal a maximum drain cur-
conductivity. rent, which is proportional to conductivity, occurring at the
To generate the maximum interaction between thio- same potential as the second polymer oxidation wave.
phene moieties and the cobalt centers we designed salen Furthermore, poly(3) shows no detectable contribution to
complexes with a para relationship between phenolic oxy- the overall conductivity from the metal-centered self-ex-
gens and the thiophenes. The monomer ligands, 2 and 5, change process, consistent with what we have observed pre-
viously in copper-based polyrotaxanes in which the copper
redox wave is well removed from the polymer redox
waves.[2a] Absolute conductivity values are difficult to cal-
± culate from our interdigitated devices, as we cannot readily
[*] Prof. T. M. Swager, R. P. Kingsborough determine what portion of the polymer lies on top of or be-
Department of Chemistry
Massachusetts Institute of Technology tween the interdigitated electrodes. Additionally, the cover-
Cambridge, MA 02139 (USA) age is not expected to be uniform since our surface cov-
[**] Funding from the Office of Naval Research is greatly appreciated. erages are much less than the interelectrode spacing.
This work made use of MRSEC Shared Facilities supported by the
National Science Foundation under Award Number DMR-9400334. Clearly the polymer films must be extremely thin between

1100 Ó WILEY-VCH Verlag GmbH, D-69469 Weinheim, 1998 0935-9648/98/1410-1100 $ 17.50+.50/0 Adv. Mater. 1998, 10, No. 14
Communications

Scheme 1. a) i) 2-Tributylstannylthiophene, cat. Cl2Pd(PPh3)2, DMF, 80 C, 12 h, Ar, 55±60 %; ii) 0.52 equivalents ethylenediamine, benzene, RT, 12 h, 94 %;
iii) Co(OAc)2×4 H2O, DMF, Ar, RT, 83 %. b) i) 2-Tributylstannyl-3,4-ethylenedioxythiophene, cat. Cl2Pd(PPh3)2, DMF, 80 C, 12 h, Ar, 27 %; ii) 0.52 equivalents
ethylenediamine, benzene, RT, 12 h, 94 %; iii) Co(OAc)2×4 H2O, DMF, Ar, RT, 83 %.

Fig. 1. Anodic polymerization of 3 on a platinum button electrode in 0.1 M

Bu4NPF6 /MeCN under a N2 atmosphere at a scan rate of 100 mV/s. Poten-
tials are referenced to the Fc/Fc+ couple.

the electrodes. Zotti and co-workers[10] have developed a

method for determining conductivities of materials by a
similar method and have used a standard material, poly(3-
methylthiophene), for which the bulk conductivity is
known (60 S cm±1),[11] to correct for non-uniform coverage.
By this method we have determined poly(3) to have a con-
ductivity of 40 S cm±1.[12,13]
From the conductivity studies it seems that although the
cobalt center is an integral part of the poly(3) backbone it
does not contribute to the overall conductivity. To confirm
the importance of the metal centers to these systems we
have attempted the oxidative polymerization of Schiff
bases based on 5-(2-thienyl)-2-methoxybenzaldehyde (7).
No film growth was observed in these systems. In fact, for Fig. 2. a) Cyclic voltammogram of a thin film of poly(3) in 0.1 M Bu4NPF6 /
this last compound the thiophene moieties are very resis- MeCN on a platinum button electrode. b) Cyclic voltammogram of poly(3)
on an interdigitated microelectrode with 5 mm spacing. c) Conductivity ver-
tant to oxidation, being oxidized at potentials similar to sus electrochemical potential of the same sample of poly(3) as in (b) on in-
thiophene itself,[14] thus emphasizing the metal-centered terdigitated microelectrodes.

Adv. Mater. 1998, 10, No. 14 Ó WILEY-VCH Verlag GmbH, D-69469 Weinheim, 1998 0935-9648/98/1410-1101 $ 17.50+.50/0 1101
 Communications

Scheme 2.

Scheme 3.

conjugation in these monomers and polymers. From the

relatively low oxidation potential displayed by 3 with re-
spect to 7, it seems that 3 behaves similar to an oligothio-
phene as regards its oxidation behavior.
In previous studies of transition metal±containing poly-
mers,[2c] we have found that using an EDOT-based (where
EDOT stands for 3,4-ethylenedioxythiophene) organic
polymer backbone leads to an overall decrease of the poly-
mer oxidation potential. Thus, we redesigned our polymer
such that the oxidation potentials of the cobalt and the or-
ganic backbone would at least partially overlap. Electro-
chemical polymerization (Fig. 3) of saturated solutions of 6

Fig. 3. Anodic polymerization of 6 on an interdigitated microelectrode with

5 mm spacing in 0.1 M Bu4NPF6 /MeCN at a scan rate of 100 mV/s.

([6] ca. 0.1 mM) immediately reveals the benefits of the

more electron-rich polymer backbone. As can be seen in Fig. 4. a) Cyclic voltammogram of poly(6) on a platinum button electrode in
0.1 M Bu4NPF6 /MeCN (solid line) and in the presence of 40 mM pyridine/
Figure 4a, the redox potential of the poly(6) backbone has
0.1 M Bu4NPF6 /MeCN (dashed line), illustrating the presence of the cobalt-
been significantly lowered to almost perfectly coincide with based redox wave. b) Conductivity versus electrochemical potential of
the Co2+/3+ redox wave. Close inspection of the CV trace poly(6) on an interdigitated microelectrode before (solid line) and after
(dashed line) exposure to pyridine.
reveals that there are two redox waves seen at ±0.11 V and
0.07 V, due to the Co2+/3+ and polymer redox processes, re-
spectively. The identity of the Co2+/3+ wave was revealed by with shifts observed in monomeric systems.[15] Also
the addition of pyridine (Fig. 4a), resulting in a shift of the observed in the cyclic voltammogram at lower potentials
wave at ±0.11 V to lower potential (±0.38 V), consistent (±0.84 V) is a wave that has been assigned to the Co+/2+ re-

1102 Ó WILEY-VCH Verlag GmbH, D-69469 Weinheim, 1998 0935-9648/98/1410-1102 $ 17.50+.50/0 Adv. Mater. 1998, 10, No. 14
Communications

dox transition. In a similar fashion to poly(3) films, it ap- in the complete, but reversible, passivation of the electrode
pears that only a small fraction of the cobalt centers can surface.[5j] Electrochemistry in the presence of the sterically
undergo this additional redox process in the thicker films. more encumbered 2,6-di-tert-butylpyridine also resulted in
The conductivity of the poly(6) film on interdigitated a small decrease (ca. 10 %) in polymer conductivity. How-
microelectrodes shows a broad trace increasing from ca. ever, unlike with lutidine, the conductivity of the polymers
±0.4 V through a maximum at 0.05 V and a subsequent de- can be restored upon rinsing with fresh electrolyte. These
crease to a high plateau at higher potentials (Fig. 4b). Con- results tend to suggest that there is a significant interaction
ductivity values[13] calculated using poly(3-methylthio- of these Lewis bases with the cobalt centers in both oxida-
phene) as a standard are quite high, ca. 250 S cm±1, roughly tion states and that the Co2+ centers remain five-coordinate
equal to that of poly(EDOT).[12] after reduction. However, at present we cannot rule out in-
The anodic polymerization of 3 was also carried out in teractions of the pyridine with the organic portion of the
0.1 M Et4NPF6 electrolyte solutions in order to determine polymer backbone.
whether cation transport through the film was affecting the In summary, we have synthesized conducting polymer±
electroactivity of the Co2+/3+ redox wave. In this case, there transition metal hybrid materials that display high conduc-
is a significant enhancement of the Co2+/3+ wave relative to tivity and a high sensitivity to coordinating ligands. We
the polymer-based waves, even in films sufficiently thick to have found that by controlling the oxidation potential of
connect the microelectrodes. This may be due to an en- the polymer relative to that of the metal center, the overall
hanced transport of the smaller cation or an overall mor- electrochemical properties of the polymer can be varied.
phological difference between the films grown in different Our future studies will seek to address the roles of the met-
electrolytes that allows all of the cobalt centers to undergo al center and the electrolyte in the conduction properties,
reversible redox processes while in the insulating regime of provide a further elucidation of the conduction mechanism,
the polymer backbone. While the exact nature of this en- as well as further the development of potential sensory and
hancement is unknown, it appears that films of poly(3) are catalytic applications.
sensitive to the electrolyte used in the film synthesis. Simi-
lar studies of the redox-matched system, poly(6), showed
no difference in the cyclic voltammograms when grown in
Experimental
either Et4NPF6 or Bu4NPF6. These results reinforce our
original premise that for these materials to have useful 5-(2-Thienyl)salicylaldehyde (1): To a mixture of 5-bromosalicylaldehyde
electrocatalytic and sensory applications, the passivation of (2.847 g, 14.16 mmol) and Cl2Pd(PPh3)2 (0.497 g, 0.708 mmol, 0.05 equiv.)
the Co2+/3+ redox wave must be overcome. Enhancement of was added tributylstannyl thiophene (6.75 mL, 21.24 mmol, 1.5 equiv.) in
25 mL of DMF. The reaction mixture was heated at 80 C for 16 h, over
the electroactivity that is observed for poly(6) in either which time the reaction color changed from light yellow to dark red. The re-
electrolyte system shows that this material's activity is not action was cooled, diluted with Et2O, and washed with dilute NH4Cl (4 ´
150 mL). The organic layer was collected and filtered through a short silica
subject to the conditions in which the polymer was synthe- pad to yield a yellow filtrate. The solvent was removed under reduced pres-
sized, as is the case for poly(3). sure to afford a gummy yellow solid, which was transferred to a frit and
Spectroelectrochemical measurements of both polymers washed with hexanes. The yellow solid was dried in vacuo to afford 1.895 g
(9.28 mmol, 66 %) of a yellow powder (m.p. 102±103 C). 1H NMR
on indium tin oxide (ITO) glass electrodes were performed (250 MHz, CDCl3) d 11.01 (s, 1H), 9.96 (s, 1H), 7.80±7.76 (m, 2H), 7.29±7.24
to elucidate the conduction mechanism. In their reduced (m, 2H), 7.09 (dd, 1H, J = 3.8 and 5.1 Hz), 7.04 (dd, 1H, J = 2.5 and 7.0 Hz).
13
forms, the polymers display broad absorptions in the visible C NMR (62 MHz, CDCl3) d 196.5, 161.1, 142.7, 134.7, 130.8, 128.3, 127.1,
124.8, 123.0, 120.8, 118.4. UV-vis (MeCN) lmax 210, 252, 289, 361 nm. MS
region with lmax = 432 nm for poly(3) and 488 nm for m/z 204 (M+). HRMS (FAB) found m/z 205.0324 (M + H+); calcd. for
poly(6), comparable to those observed by Goldsby in poly- C11H9O2S m/z 205.0323 ([M + H]+). Elemental anal. calcd. for C11H9O2S:
C, 64.69; H, 3.95. Found: C, 64.76; H, 3.78.
mers of Cu(salen) and Ni(salen).[5a] As the polymers are 5-(2-(3,4-Ethylenedioxy)thienyl)salicylaldehyde (4): This compound was
oxidized, this absorbance shifts to higher energy and rapid- prepared in a similar fashion to 1 except that 2-tributylstannyl-3,4-ethylene-
ly decreases in intensity. Concurrent with this shift is the dioxythiophene was used in place of thiophene (27 %). m.p. 132 C, dec. 1H
NMR (250 MHz, CDCl3) d 10.98 (s, 1H), 9.94 (s, 1H), 7.92 (d, 1H, J =
formation of two new peaks at lower energy, which extend 2.3 Hz), 7.84 (dd, 1H, J = 2.4 and 8.7 Hz), 7.00 (d, 1H, J = 8.7 Hz), 6.29 (s,
into the near IR and are often considered to be the signa- 1H), 4.34±4.25 (m, 4H). 13C NMR (125 MHz, CDCl3) d 197.1, 160.4, 142.6,
ture of free carriers in the metallic state.[16] This type of 138.2, 134.9, 131.1, 125.9, 120.8, 118.2, 115.8, 97.5, 65.1, 64.7. UV-vis (MeCN)
lmax 214, 257, 295, 373 nm. MS m/z 262 (M+). HRMS (FAB) found m/z
behavior has also been observed in purely organic conduct- 263.0378 (M + H+); calcd. for C13H11O4S m/z 263.0378 ([M + H]+). Elemen-
ing polymer systems.[9c,9d,16,17] tal anal. calcd. for C13H11O4S: C, 59.53; H, 3.84. Found: C, 59.35; H, 3.86.
The ligands [6a] and metal complexes [6b] were prepared by literature
Exposure of the polymers to pyridine (ca. 40 mM) results procedures from the appropriate salicylaldehyde.
in an irreversible, approximately 66 % quenching of the N,N¢-Ethylenebis(5-(2-thienyl)salicylidenimine) (2): 94 % yellow powder,
conductivity of poly(6) (Fig. 4b). Rinsing the polymer with m.p. 255 C, dec. 1H NMR (250 MHz, CDCl3) d 13.29 (s, 2H), 8.43 (s, 2H),
7.55 (dd, 2H, J = 2.4 and 8.4 Hz), 7.46 (d, 2H, J = 2.3 Hz), 7.22±7.16 (m,
fresh electrolyte did not result in a restoration of the origi- 4H), 7.05 (dd, 2H, J = 3.6 and 5.0 Hz), 6.97 (d, 2H, J = 8.7 Hz), 3.99 (s, 4H).
nal conductivity. Addition of 2,6-lutidine also results in the The poor solubility of this compound prevented characterization by 13C
same film behavior. While no conductivity studies have NMR spectroscopy. UV-vis (MeCN) lmax 252, 293 nm. MS m/z 432 (M+).
HRMS (EI) found m/z 432.0964 (M+); calcd. for C24H20N2O2S2 (M+)
been performed, others have observed that exposure of 432.0966. Elemental anal. calcd. for C24H20N2O2S2: C, 66.64; H, 4.66; N,
pyridine to transition metal±based polymer systems results 6.48. Found: 66.43; H, 4.35; N, 6.34.

Adv. Mater. 1998, 10, No. 14 Ó WILEY-VCH Verlag GmbH, D-69469 Weinheim, 1998 0935-9648/98/1410-1103 $ 17.50+.50/0 1103
 Communications

N,N¢-Ethylenebis(5-(2-(3,4-ethylenedioxy)thienyl)salicylidenimine) (5): the polymer thickness; with a 5 mm gap, D; n = 99 gaps; and a

94 % yellow solid, m.p. 245 C, dec. 1H NMR (250 MHz, CDCl3) d 13.26 (s, 0.49895 cm electrode length, L.
2H), 8.41 (s, 2H), 7.63±7.60 (m, 4H), 6.95 (d, 2H, J = 8.3 Hz), 6.22 (s, 2H),
iD D
4.35±4.22 (m, 8H), 3.97 (s, 4H). The poor solubility of this compound pre- sˆ  (1)
V D n
T
L
vented characterization by 13C NMR spectroscopy. MS m/z 549 ([M + H]+).
HRMS (FAB) found m/z 549.1157 ([M + H]+); calcd. for C28H25N2O6S2 m/z
[14] J. Roncali, Chem. Rev. 1992, 92, 711.
549.1154 ([M + H]+).
[15] E. Eichhorn, A. Rieker, B. Speiser, H. Stahl, Inorg. Chem. 1997, 36,
N,N¢-Ethylenebis(5-(2-thienyl)salicylideniminato)cobalt(II) (3): 83 %
3307.
black microcrystalline solid, m.p. >250 C. UV-vis (MeCN) lmax 261, 315,
[16] a) A. O. Patil, A. J. Heeger, F. Wudl, Chem. Rev. 1988, 88, 183. b) X.
389, 415 nm. MS m/z 489 (M+). HRMS (EI) found m/z 489.0144 (M+); calcd.
Chen. O. Inganäs, J. Phys. Chem. 1996, 100, 15 202.
for C24H18CoN2O2S2 (M+) 489.0142. Elemental anal. calcd. for C24H18Co-
[17] See, for example: a) G. A. Sotzing, J. L. Reddinger, A. R. Katritzky, J.
N2O2S2: C, 58.89; H, 3.71; N, 5.72. Found: C, 58.76; H, 3.94; N, 6.11.
Soloducho, R. Musgrave, J. R. Reynolds, Chem. Mater. 1997, 9, 1578.
N,N¢-Ethylenebis(5-(2-(3,4-ethylenedioxy)thienyl)salicylideniminato)co-
b) G. A. Sotzing, J. R. Reynolds, P. J. Steel, Chem. Mater. 1996, 8, 882.
balt(II) (6): 83 % green-black microcrystalline solid, m.p. >250 C. UV-vis
(MeCN) lmax 267, 317, 398 nm. HRMS (LDMS) found m/z 605.02 (M+);
calcd. for C24H18CoN2O2S2 (M+) 605.54. Elemental anal. calcd. for
C28H22CoN2O6S2: C, 55.54; H, 3.66; N, 4.63. Found: C, 54.73; H, 3.67; N,
4.56.
Superlattice Structures in
Poly(phenylenevinylene)-Based
Received: April 7, 1998
Final version: June 4, 1998 Self-Assembled Films**
±
[1] a) Handbook of Conducting Polymers, 2nd ed. (Eds: T. A. Skotheim, By Haiping Hong, Roland Steitz, Stefan Kirstein, and
R. L. Elsenbaumer, J. R. Reynolds), Marcel Dekker, New York 1998.
Dan Davidov*
b) P. Audebert, Curr. Top. Electrochem. 1994, 3, 459. c) D. M. Kelley,
J. G. Vos, in Electroactive Polymer Electrochemistry (Ed: M. E. Lyons),
Plenum, New York 1994, Vol. 2. d) M. Lyons, in Electroactive Polymer Several years ago, Decher and co-workers presented a
Electrochemistry (Ed: M. E. Lyons), Plenum, New York 1994, Vol. 1.
[2] a) S. S. Zhu, T. M. Swager, Adv. Mater. 1996, 8, 497. b) S. S. Zhu, P. J.
new layer-by-layer self-assembly method to fabricate poly-
Carroll, T. M. Swager, J. Am. Chem. Soc. 1996, 118, 8713. c) S. S. Zhu, meric films consisting of layers of different polymers.[1,2]
T. M. Swager, J. Am. Chem. Soc. 1997, 119, 12 568. These multilayer structures are prepared by alternate dip-
[3] a) B. S. Tovrog, R. S. Drago, J. Am. Chem. Soc. 1974, 96, 6765.
b) B. M. Hoffman, T. Szymanski, F. Basolo, J. Am. Chem. Soc. 1975, ping of a charged substrate into an aqueous solution of a
97, 637. positively charged polyelectrolyte and another solution of
[4] a) R. Cini, P. L. Orioli, J. Chem. Soc., Chem. Commun. 1981, 196. b) P.
a negatively charged polyelectrolyte; the adsorption of al-
Zanello, R. Cini, A. Cinquantini, P. L. Orioli, J. Chem. Soc., Chem.
Commun. 1983, 2159. ternate layers is driven by electrostatic interactions. In this
[5] See, for example: a) L. A. Hoferkamp, K. A. Goldsby, Chem. Mater. way layered multicomposites with multifunctional proper-
1989, 1, 348. b) K. A. Goldsby, J. K. Blaho, J. K. Hoferkamp, Polyhe-
dron 1989, 8, 113. c) K. A. Goldsby, J. Coord. Chem. 1988, 19, 3117.
ties can be built, which are of interest for various electro-
d) P. Audebert, P. Capdeville, M. Maumy, New J. Chem. 1992, 16, optical applications. The method was applied successfully
697. e) P. Audebert, P. Capdeville, M. Maumy, Synth. Met. 1991, 41± to fabricate light-emitting devices (LEDs) in which the
43, 3049. f) P. Audebert, P. Capdeville, M. Maumy, New J. Chem.
1991, 15, 235. g) C. E. Dahm, D. G. Peters, Anal. Chem. 1994, 66, emitting layer is derived from the semiconducting conju-
3117. h) J. L. Reddinger, J. R. Reynolds, Synth. Met. 1997, 84, 225. i) gated polymer poly(p-phenylenevinylene) (PPV).[3±7] The
J. L. Reddinger, J. R. Reynolds, Macromolecules 1997, 30, 673. j) J. L.
fabrication of mulifunctional devices requires highly or-
Reddinger, J. R. Reynolds, Chem. Mater. 1998, 10, 3. k) J. L. Reddin-
ger, J. R. Reynolds, Chem. Mater. 1998, 10. 1236. l) F. Bedioui, E. dered and stable multilayer systems in which the different
Labbe, S. Gutierrez-Grandanos, J. Devynck, J. Electroanal. Chem. layers have different functionalities. It is important, there-
1991, 3011, 267.
[6] a) D. J. Aymes, M. R. Paris, J. Chem. Educ. 1989, 66, 854. b) C. Floria-
fore, to control and minimize the interpenetration of the
ni, F. Calderazzo, J. Chem. Soc. 1969, 946. different polymers at the interfaces (buried interfaces) such
[7] Electrochemical studies were performed under a nitrogen atmosphere that the polymer penetration depth will not exceed 1±
in 0.1 M Bu4NPF6 /MeCN electrolyte solution on either a platinum
button electrode (A = 0.02 cm2) or 5 mm interdigitated microelec- 2 nm. This interpenetration depth at the interface is di-
trodes (A = 0.0496 cm2) with a platinum coil auxiliary electrode and an rectly related to the interlayer roughness.
isolated silver/silver nitrate reference electrode using an EcoChemie
Autolab bipotentiostat. All electrochemical data are referred to ferro-
cene/ferrocenium (Fc/Fc+). ±
[8] C. P. Horwitz, R. W. Murray, Mol. Cryst. Liq. Cryst. 1988, 160, 389. [*] Prof. D. Davidov, H. Hong
[9] a) G. P. Kittlesen, H. S. White, M. S. Wrighton, J. Am. Chem. Soc. Racah Institute of Physics
1984, 106, 7389. b) D. Ofer, R. M. Crooks, M. S. Wighton, J. Am. The Hebrew University of Jerusalem
Chem. Soc. 1990, 112, 7869. c) E. W. Paul, A. J. Ricco, M. S. Wrighton, Jerusalem 91904 (Israel)
J. Phys. Chem. 1985, 89, 1441. d) J. W. Thackeray, H. S. White, M. S.
Dr. R. Steitz
Wrighton, J. Phys. Chem. 1985, 89, 5133.
Berlin Neutron Scattering Center
[10] a) G. Zotti, G. Schiavon, Synth. Met. 1990, 39, 183. b) G. Schiavon, S.
Hahn-Meitner-Institut
Sitran, G. Zotti, Synth. Met. 1989, 32, 209.
D-14109 Berlin (Germany) and
[11] G. Tourillon, G. Garnier, J. Electroanal. Chem. 1982, 135, 173.
Iwan N. Stranski Institut
[12] As an independent check of Zotti's method with our devices, we deter-
Techniche Universität Berlin
mined the conductivity value of poly(3,4-ethylenedioxythiophene)
D-10623 Berlin (Germany)
(poly(EDOT)) to be ca. 110 S cm±1, only a factor of two lower than
that reported in the literature for electrochemically grown material: G. Dr. S. Kirstein
Heywang, F. Jonas, Adv. Mater. 1992, 4, 116. Max-Planck-Institute fur Kolloid- und Grenzflachenforschung
[13] Film thickness measurements were performed on a Tencor P10 Sur- Rudower Chausee 5, D-12489 Berlin (Germany)
face Profilometer. The minimum thickness was used to calculate the [**] This work was supported by the VW foundation, the Berlin Neutron
conductivity (s) of the film. Conductivity was calculated using Equa- Scattering Center (BENSC), and the Israeli Ministry of Science. We
tion 1, where iD is the drain current, VD is the offset potential, and T is are grateful to M. Tarabia for help with the fitting procedure.

1104 Ó WILEY-VCH Verlag GmbH, D-69469 Weinheim, 1998 0935-9648/98/1410-1104 $ 17.50+.50/0 Adv. Mater. 1998, 10, No. 14