Mol. Cry". Liq. Cryst., 1982, Vol. 83, pp. 217-227002~8941182/8301-0217$06. 50/0e ~982Gordon and Breach, Science Publishers IncPrinted in the United States of America "(cPrOceedingsof the International Conference on Low_Dimensional Conductors, Boulder,

olorado, August 1981)

CONDUCTINGCOMPLEXESOF CONJUGATEDPOLYMERS:

A COMPARATIVESTUDY

R.R. CHANCE,R.H. BAUGHMAN,J.L. BREDAS,t H. ECKHARDT,R.L. ELSENBAUMER,J.E. FROMMER,and L.W. SHACKLETTE

Polymer LaboratoryAllied Corporation,Morristown, N.J. 07960

R. SILBEYDepartment of ChemistryMASSACHUSETTSINSTITUTE OF TECHNOLOGY,Cambridge, MA02139

Submitted for publication August 31, 1981

There are now a number of different polymerS which under-go dramatic conductivity increases on exposure to electrondonors or acceptors. The discussion here will center on acomparison of the electrical and optical properties of thedifferent systems, with special emphasis on polyphenylenesand polyphenylenesulfides. The very promising resultsprovided by the Valence Effective Hamiltonian (VEH) tech-nique for the valence band electronic structures of a widerange of undoped polymerS will provide the framework for acomparison of the various conducting polymer systems.

r 1. INTRODUCTIONConjugated polymers have for some time been of major interestto chemists and physicists alike. This interest has been en-hanced greatly in recent years by the discoveries of a numberof conjugated polymer systemsl-4 which could be "doped" tonear metallic conductivities. The doping process involvesexposure of the polymer to electron donors, such as the

tFellow of the Belgian National Science Foundation

(Aspirant FRS)

[1249]1217

218/[1250] R. R. CHANCE et al.

alkali metals, or electron acceptors, such as 12 and AsFSoDoping can also be accomplished electrochemically _ a factwhich has lead to rechargeable batteries based on conductingpolymers.S Doping levels are generally quite high (typicallyS-SO mole per cent) so that the phrase "doped polymers" can tea bi t misleading if taken in the context of semiconductor phy- Isics. For doped inorganic semiconductors, the dopant isusually at very low concentrations and it represents a rela-tively minor perturbation on the band structure of the saaicon-ductor and, therefore, on transport properties. This rigid bandstructure model is of limited applicability, even qualitatively"to doped conjugated polymers. For this reason, as well as thefact that many aspects of the electronic properties of thesematerials are not well-established theoretical understanding

, i e~of conducting polymers lags behind the rapid pace of exper mtal developments. This statement includes theory related to

COO"transport properties and theory related to predicting newconducting systems.In this paper. we will briefly summarize the optical and

electrical properties of phenylene-based conducting polymerd cep'systems: donor and acceptor doped poly(p-phenylene) an ac

tor doped poly(p-phenylene sulfide). Comparisons will be madet·calto the much studied polyacetylene system. Recent theore 1

k6 i d . f r acCOP-wor a me at developing SOmepredict1Ve capability 0tor doped systems will be discussed.

II. ELECTRICALAND OPTICALPROPERTIES

Several conducting polymer systems have been discoveredrecently including: poly(p-phenylene), PPP, announced byIvory ~ al,2 and poly(p-phenylene sulfide), PPS, announcedsimultaneously by Rabolt et al.3 and Chance et al.4 The PPSsystem is especially signUicant since it represents thefirst conducting system Containi~g a polymer which is pro-cessible by conventional polymer techniques.]

Table I summarizes maximumconductivities achieved forpolyacetylene (PA), PPP, and PPS on doping with 12 (a rela-tively weak electron acceptor) AsFS (a strong electron I'at) , I orksccep or , and potassium (a strong electron donor). 2 W Ponly for PA, AsFS for all three, and potassium for PA and PP •In each case the sign of the carriers as measured by Hall teffect, thermopower, etc.,1-4,7-8 is ~onsistent with transporin the polymer array, i.e., p-type for acceptor doping andn-type for donor doping.

CONDUCTING COMPLEXES OF CONJUGATED POLYMERS [1251]/219

lable I. Conductivities (S/cm) for doped polymersa

Polymer 0[12] a[AsFs] a[K]b

PA~ X550 [0.13] 1200 [0.20] 50 [0.32]

PPP-f@-1:X

<10-4 500 [0.40] 20 [0.50]

PPS@-S+X

aTypical d iprob bl op ng levels are given in bra.\'kets on the basis ofas iadie dopant species (13-, AsF6-, K) per monomeric unitb n cated in the first column.

Results obtained byth 1 exposure of polymer to potassium naph-

c a ide solution (THY).

Doping. at this level is accompanied by extensive chemicalmodlficatibenzothi on involving intramolecular crosslinking to form

ophene linkages.7

l Important fsyst actors in determining whether or not a conductingem is fpotentia ormed on acceptor doping are: (L) the ionization

be d 1 1 of the polymer and (2) the extent to ",hich charge canor e ocalized along the chain The former determines ",hether

not ch •dete arge-transfer occurs with a given acceptor; the latterrmines h d hchai '" ether or not the charges (holes) induce in t e

• n will b . 11,"qui e mobile. Thus conJugated polymerS are genera Yto 1 red since electron delocalization in these systems leads

ow ionization potentials and large ._electron band",idths.

220/[1252] R. R. CHANCE et al

Conjugation in IT-electron systems is generally assoctsteiwith planarity, which affords maximumn-electron overlap.Thus, at first glance, it is surprising that PPP and PPSform conducting complexes since both are nonplanar systems.Adjacent phenyl rings in PPP are rotated by _23' with respectto one another. Adjacent phenyl rings in PPS are nearly per'pendicular. In addition, the presence of aromatic moietiesmight be expected to lead to localization, since, for ,example, in the PPP case a loss of aromaticity is required In

forming the mesomeric quinoid structure. We return to some ofthese questions in Section III. However, it is clear fromTable I that PPP and PPS have higher ionization potentialsthan PA. The data also suggest that PPS has a lower electronaffinity than either PPP or PA.

An important effect in the PPP and PPS systems is thepotential for chemical modification on doping. Shacklette~ ~9 have shown that oligomers of PPP (biphenyl,P'rt.er pheny l , etc.) polymerize to form PPP on exposure toAsFS, eventually forming a conducting PPP/AsFS complex withconductivity comparable to that obtained starting with thepolymer. In the PPS case, exposure to high levels of AsFSorSbFS results in substantial intramolecular bridging to form

tras·dibenzothiophene linkages as indicated by infrared spec'7 ipwcopy and chemical analyses. The limit of this bridg ng

cess is polybenzothiophene. The highest conductivity"achieved in PPS Without measurable chemical modification is-0.01 S/cm.7

The optical properties of the polymers undergo dramaticchanges on doping. Electronic absorptions intrinsic to thepolymer decrease in intensity and are replaced by neWtransitions in the near-infrared region. 4,9-11 At high x'doping levels, the absorption due to the new transitions e dtends to very low energies effectively masking the infrareVibrational transitions Of' the polymer. Spectroscopic re-sults are summarized in Fig 1 for PA 10 PPP 9 ppS,4,7 andpolypyrole.

11Absorption b:nds intri~sic to' these polymers

are located at 1.8 eV, 3.5 eV, and 3.6 eV and 2.8 eV re-hespectively. At low doping levels, new peaks appear in t

near-infrared (around 1 eV) for PA, PPS, and polypyrole•For PPP, spectra at low doping levels have not, as yet, 'n-been obtained. The spectrum for PPP in Figure 1 was obta1 sed after expos Log a terphenYl film to AsFS which polymerize~he oligo

mersand forms PPP/AsFS complexes:9 At high doping

evels, the spectra for PAlAsFS show absorption through- hout the infrared (as for PPP). For PPS the spectra at higdoping levels do not show nearly as much absorption in theitbinfrared. This weaker infrared absorption is consistent ~the lower conductivity for PPS/AsF

5.

CONDUCTING COMPLEXES OF CONJUGATED POLYMERS [1253]/221

-@-s-

H HC-c• •-c c-'N'H

2 3

PhotonEnergy (eVI

Absorption spectra for acceptor-doped polymers

For PA a hput fo ' t eoretical model based on solitons has beenrward to 1transiti exp ain the dopant-induced near-infrared

of FIGU~n1at 0.8 eV.12 However, since the other polymersall-trans PAlack the two degenerate ground states present insystemati ,they cannot support soliton states, and aSoliton cdexplanation of all the FIGURE1 data with aohere thmo el is not possible. Br"das et al

l3have shown eise-

at a pol d --qualitat' aron mo el offers an internally consistent,near-inf ive description of these data. In that model, the

rared abs b' d di 1cation (. or ang species is a delocalize ra ca.r-'electro 1. e , , a hole or polaron) produced by loss of anof delon ~o the acceptor. This model suggests that the extentPhenYlca

iization of the polaron in PPP is about three or four

r ngs.The ext:pparentl ent of delocalization in the PPS systems is also

doping" y about 3 or 4 monomeric units, as judged from the("tetrame~: PPS "oligomers": (C6H4)3S2 ("trimer") or (C6

H4)4S3

Onexpo ). The near infrared transition energies, inducedSure t A Fare 1.55 lOS 5. of the trimer. tetramer, and polymer

tetramer' .24, and 1.21 eV respectively. Thus theregard t and the polymer are nearly indistinguishable withthe do 0 optical properties However the conductivity of

ped oligomers is less ~han 10-7 S/cm.

FIGURE1

222/[1254] R. R. CHANCEet al.

III. VEHCALCULATIONS

It has been demonstrated that good insight into the groundstate properties of polymeric sl;stems can be obtained byquantum chemical calculations. 1 However, the SCF (Self-Consistent-Field) ab initio Hartree-Fock techniques, whichhave proved very successful on small molecules, becomealmostprohibitively expensive when applied to polymers of interest,Cheaper semiempirical techniques could in principle be used,but are generally less reliable.

Recently, Nicolas and Durand15 have developed an inter-esting approach based on the use of valence effective 10Hamiltonians (VEH). Applied to hydrocarbon molecules, theVEHtechnique affords one-electron energy levels of ab inltiodouble zeta quality at a cost comparable to semiempiricaltechniques such as Extended Hucker. The VEHmodel has re-cently been applied to polymer calculations. 7,16 It has beeoshown,7 through calculations on polyacetylene and polydiacet-lyene, that the polymer VEHmethod also provides ab initiodouble zeta quality results for band structure and densityof states. Perhaps more important with regard to the con-ducting polymers area, the VEHtechnique provides reliableionization potentials and bandwidths. The ionizationpotential determines whether a particular acceptor isof ionizing (or partially ionizing) the polymer. Thewidth of the highest occupied band provides a measureextent of the delocalization in the system and can be

riersroughly correlated with the mobilities of the charge car din that band. Thus bandwidths of the highest occupied banshould show a qualitative correlation to conductivitiesachieved in similar polymers upon acceptor doping. Cautionis needed on the latter point because of the demonstratedinadequacy of a rigid band model to explain the propertiesof highly conducting doped polymers. 17

In this section we report in summary form, VEHcalculations for a wide range' of hydrocarbon polymers inorder to provide a systematic and theoretically consistentdescription of the valence electronic properties of systemsof interest in the conducting polymers area The completemethodology for obtaining molecular one-ele~tron effectiveHamiltonians from first principles has been developed inrei. 15 and has been extended for polymer sys terns in e'refs. 7 and 16. The main advantages of the VEHmodel ar .(L) it is completely theoretical; (11) it is not basis S~~tbdependent; and (11i) it gives ab initio quality results ne-negligi bLe computer time due to the evaluation of only 0 i"electron integrals and the complete avoidance of SCFiteratcycles.

capableband-of the

CONDUCTINGCOMPLEXESOF CONJUGATEDPOLYMERS[1255]1223

The VEHresults for ionization potential (IP) bandwidthof the highest occupied band (BW), and .-electron bandgap(Eg)are summarized in Table II. Also included in the tableare experimental estimates of IP and E where available andthe maximumconductivities (0) obtaine§ on doping with 12 (arelatively weak electron acceptor) and AsFS (a strong electronacceptor). Representati ve references for the experimental IP,Egoand a values are i-s, 7-11, 18-27. The theoretical IPvalues have been scaled downward by 1.9 eV to correct approxi-mately for polarization energy and possible shortcomings ofthe model.7 Though the 1.9 eV correction is quitereasonable,28 our IP results should be viewed as having beenscaled to the experimental estimate for trans polyacety-lena, 8,18The IP values are in good agreement with available experi-

mentalestimates. The IP for PPP and PPS are significantlylarger than PA, thus explaining why 12 works only for PA. TheIP for polybenzothiophene, PBT, the product expected after:ntramolecular bridging of PPS, is seen to have a substantial-y lower IP than PPS. In fact, our theoretical estimate agreeswith the experimental estimate from junction measurements onheaVily doped (and therefore chemically altered) PPS. The 31values also show a satisfactory qualitative correlation to theconductivities achieved on AsFS doping.

Note that the BW value for PPS is fairly large, 1.2 eV,comparedto the perpendicular PPP case 0.2 eV. This resultIs d 'ha irect indication of the important role played by t esulfur atoms in providing a communication path for the'-systems of the perpendicularly oriented phenyl rings in PPS.

The Eg values obtained by the VEH technique are insurprisingly good agreement with experiment - especially when~e consider the fact that no excited state information haskeen included in the parameterization because of the well-nowu poor performance of one-electron ab initio techniques

, regarding excited state energies. Eg values for nonplanar?stems are not given because of the previously mentio

ned6

endency f 1 I 0* bands.or this method to give spuriouS Y ow~(R)Thetheoretical results for polydiacetylenes,hi hi-C=C-C(R)F, suggest these materials should also for:

hg Y conducting complexes perhaps even with iodine. eaVed d ' R roupS) withlope a number of PDA systems (different g2 and A F h re a conductivitygre t s S, but have not found any case wed i gs bs er than 10-6 8/cm can be achieved. With AsFS op n ,I~d~tantial chemical interaction with the R groupS is fth c:ated in Some cases The large R groupS, necessary 1 orImeSolid-state Synthes~s of PDAsingle crystals, may a sOlPose st f rable polymerdo eric restrictions which inhibit avOpant interaction.

224/[1256] R. R. CHANCE et af.

TABLEII. Summary of VEH results for ionization potential (IP),a bandwidthof highest occupied band (BW), and bandgap (Eg). All energiesIn eV; c.onduc.tivities In Stem. b

Polymer 1po Il\Id E 0 a(AsF5) 0(12)g

Polyac.etylenetrans 4.7 14.7] 6.5 1.4 [1.8] 1200 500c!s-traosold 4.8 6.4 1.5 [2.0)trans-dsoid 4.7 6.5 1.3

Polymethylacetylene 4.5 3.7 1.4 )0"3

Poly(1,6-heptadiyne) 4.4 2.5 1.4 [1.8) O. I 0.1

Polydiacetyleneacetylenlc 5.1 [5.2J 3.9 2.1 [2.1) 0 0butatrienic. 4.3 4.5

Poly(p-phenylene)coplanar 5.5 3.9 3.2twisted (220) 5.6 [5.5] 3.5 [3.5) 500 0perpendicular 6.9 0.2

Poly(m-phenylene)coplanar 6.2 0.7 4.5twisted (28°) 6.2 0.2 [4.9] 0.001 0

Poly(p-phenylenevlnylene) 5.1 2.8 2.5 [-3) 3 0

Poly(p-phenylenexylylidene) 5.6 2.5 3.4

Poly benzyl 6.5 0.6 0 0Poly(p-phenylene 6.3 1.2 0.01 0sulfide)

POlybenzothlophene 5.5 [5.6] 1.3 3.1 [-3J 3

aTheoretical IP after sUbtracting

1.9 eV to correct, approximately, forbPolarization energy.

Zero conductiVity in the table i di t kno\l/ll'c - n ca es 0<10-5 S/cm; dash indicateS unExperimental estimates

for IP and Ego where available, aregiven in bracketad •

BWvalues refer to thesmallest possible unit cell to''''·axis sYIIlIlletry. taking into acc.oLl"

CONDUCTINGCOMPLEXESOF CONJUGATEDPOLYMERS[1257]/225

IV. SUMMARY

ThoughPPP clearly has a higher ionization potential and ahigher bandgap than PA, the profiles of their optical andelectrical properties for the doped systems are very similar.Theyboth would appear to form "simple complexes" on doping.Thisbehavior is in contrast to the PPS system where chemicalmodification is seen to play an important role in achievinghighconductivity on doping. Polymers which undergo chemicalmodification may be the most technologically significant ofthe conducting polymer systems, since processible polymerssuchas PPS can be used in that case. With this dopant-inducedchemicalmodification, there need not be any conflict betweenthe flexible backbone needed for processibility and the rigidbackbonewhich is apparently required for high conductivity.Neither PPP nor PA is processible via conventional polymertechnfq ues ,The general similarity of doped-PPP and doped-PA,

particularly with regard to their electrical properties, isanobservation which requires some emphasis, especially withregard to developing theoretical models for these highlyC~~ductingdoped systems. Models, such as the soliton model,W Ich can be applied to PA but cannot be generalized to PPP,mustbe viewed with caution particularly when extended todescrib ht 'e 19h doping levels.

V. ACKN<MLEDGEMENTS

This kF war was supported in part byoundation (Grant II DMR-7908307).

the National Science

226/[1258J R. R. CHANCE et al.

REFERENCES

1. C.K. Chiang, C.R. Fincher, Y.W. Park, A.J. Heeger,H. Shirakawa, E.J. Louis, S.C. Gau, and A.G. MacDiarmid,Phys. Rev. Lett.39, 1098 (1977).

2. D.M. Ivory, G.G.~iller, J.M. Sowa, L.W. Shacklette,R.R. Chance, and R.H. Baughman, J. Chem. Phys. 71,1506 (1979); L.W. Shacklette, R.R. Chance, D.M. Ivory,G.G. Miller, and R.H. Baughman, Synthetic Metals l, 307(1980).

3. J.F. Rabolt, K.K. Kanazawa, J.R. Reynolds, and G.B.Street, J. Chem, Soc. Chem, Comm, 347. (1980).

4. R.R. Chance, L.W. Shacklette, G.G. Miller, D.M. Ivory,J.M. Sowa, R.L. Elsenbaumer, and R.H. Baughman,J. Chern, Soc. Chern, Comm, 348 (1980). d

5. P.J. Nigrey, D. MacInnes, D.P. Nairns, A.G. MacDiarmi,and A.J. Heeger, J. Electrochem. Soc_, in press

6. J.L. Bredas, R.R. Chance, R. Silbey, G. Nicolas, andPh. Durand, J. Chem. Phys , 75, 255 (1981). h dt

7 L W Sh k - h HEck ar ,• •• ac lette, R.L. Elsenbaumer, R.R. C ance , • 1919J.E. Frommer, and R.H. Baughman, J. Chem, Phy s , 71, d(1981); R.L. Elsenbaumer, L.W. Shacklette, J.M. Sowa, anR.H. Baughman in Proceedings of International Conferenceon Low-Dimensional Conductors (Boulder, August, i981).

8. T. Tani, W.D. Gill, P.M. Grant, T.C. Clarke, andG.B. Street, Synthetic Metals 1 301 (1980).

9 "ller• L.W. Shacklette, H. Eckhardt, R.R. Chance, G.G. Ml 98'D.M. Ivory, and R. H. Baughman J. Chern, Phy s , fl' 40 ,(1980). '

10.C.R. FinCher, D.L. Peebles, A.J. Heeger, M.A. Druy,Y. Matsumura, A.G. MacDiarmid, H. Shirakawa, andS. Ikeda, Solid State Comm. 27 489 (1978). t

11.K. Kanazawa, A. Diaz, W.D. Gili, P.M. Grant, G.B. Stree, )G.P. Gardini, and J.L Kwak Synthetic Metals ..!., 329(1980). '

12.M.J. Rice, Phys, Lett. A, 71 152 (1979); M.J. Rice andJ. Timonen, Phyg , Lett. A, 73 368 (1979); w. P. su, 42J.R. Schrieffer, and Ad. Heeger Phys. Rev. Lett., '1698 (1979); Phy s , Rev. B, 22 2209 (1980). s nf

13.J.L. Bredas, R.R. Chance, and R. Silbey, in proceed~~:International Conference on Low-Dimensional Conduct(Boulder, August 1981)

14. Rece t Ad' J.M.n vances in the Quantum Theory of polymerSL d

~ndre, J.L. Bredas, J. Delhalle, J. Ladik, G. LerOy/nn),(S Moser, eds., Lecture Notes in Physics Series, vo .Springer-Verlag, Berlin, 1980).

CONDUCTING COMPLEXES OF CONJUGATED POLYMERS [1259]1227

REFERENCES

15.G. Nicolas and Ph. Durand, J. Chera, Phys•• 70, 2020(919); ibid •• 72, 453 (1980). -

16.J.M. Andre, L. A-.-Burke, J. DeLhaLl.e , G. Nicolas, and Ph.Durand. Int. J. Quantum chem•• S13. 283 (1979).

17.J.L. Bredas, R.R. Chance. and R.:J. Silbey. J. Phys , Chern,(Let.t, ) 85. 756 (1981).

IS.W.R. SaLaneck , H.R. Thomas. C.B. Duke. A. Paton, E.W.Plummer. A.J. Heege r , 'and A.G. MacDiarmid. J. Chem, Phye, ,71, 2044 (1979).

19.R."Shirakawa and S. Ikeda. PoLm, J •• 2. 231 (1971)20.Y.w. Park. A.J. Heeger, M.A. Druy, and A.G. MacDiarmid.J. Chem, Phy s , , 73. 946 (1980).

21.F.E. Karasz , J.C:W. Chien. and G.E. Wnek. Org- Coat.Pl.as t , Chem•• 43. 879 (1980).

22.G.E. Wnek, J. C-.-Chien, and F. E. Karasz. urg. Coat.Pl.ast , Chem., 43, 882 (1980).

23.J.M. Pochan andH.W. Gibson. Org. Coat. Plast. Chem•• ~,872 (1980).

24.R.R. Chance. R.H. Baughman. H. Muller, and C.J. Eckhardt.J. Chem. Phys •• 67. 3616 (1977).

25.J.C.W. Chien. R.r;-:- Gooding. F.E. Karasz. C.P. Lillya. G.E.2 Wnek. and K. Yao, org , Coat. PIast. Chem., ~, 886 (1980).6.H. Eckhardt. R.H. Baughman, J.L. Bredas. R.R. Chance.R.L. Elsenbaumer and L.W. Shacklette, Mat. Sci. 7.... 121(1981). •

i7.J. Boon and E.P. Magre. Makromol. Chem•• 126, 130 (1969).S.N. Sato. K. Seki, and H. Inokuchi in 9th MolecularCrystal Sym•• 218 (1980).