Role of Buckminsterfullerene, C60, in organic photoelectric devices

Quanr. E/ec!r. 1995, Vol. 19, pp. 131-159 0 1995. Elsevier Science Ltd Printed in Great Britain. All rights reserved.

frog.

Pergamon

0079-6727(94)00012-3

0079-6727195 629.00

ROLE OF BUCKMINSTERFULLERENE, C,, ORGANIC PHOTOELECTRIC DEVICES

IN

N. SERDAR SARICIFTCI Institute of Polymers and Organic Solids, University of California, Santa Barbara, CA 93106, U.S.A. Abstract-In this review we discuss the photophysical properties of the supramolecular composites of two n-electron semiconductors; conjugated polymers as electron donors and Buckminsterfullerene as electron acceptor. Conjugated, polymeric semiconductors have been found to be effective donors upon photoexcitation of the valence band electrons across the bandgap into the conduction bands. The Buckminsterfullerene, C,, acts as a strong acceptor upon photoexcitation. Thus, the supramolecular composite of these two conjugated materials exhibit ultrafast, reversible, metastable photoinduced electron transfer and charge separation. This process, similar to the primary steps of photosynthesis, has been utilized in conjugated polymer/C, based heterojunction as well as in met&semiconductor-metal tunnel devices for effective conversion of the solar photon energy into electricity. Other related applications of the photophysics, including photolithographic and xerographic processes, are reviewed. Furthermore, recent developments on light emitting diodes and solar cells with anomalous behavior based on Buckminsterfullerene are critically reviewed. Quantum well-like heterostructures based on organic donor-acceptor layers are proposed to exhibit interesting photoinduced phenomena.

CONTENTS Abbreviations Foreword 1. Introduction 1.1. Conjugated polymeric materials: n-electron semiconductors 1.2. Relation between molecular orbital overlap and transport properties 1.3. Why a-electrons? 2. Photoinduced Electron Transfer from Conjugated Polymers onto Buckminsterfullerene, C, 2.1. Background 2.2. Direct experimental evidence for the metastable charge separation: Light induced Electron Spin Resonance (LESR) 2.3. Direct observation of 300 femtosecond forward electron transfer 2.4. Sensitization of photoconductivity 3. Buckminsterfullerene, C, , in Organic Electronic Devices 3.1. Conjugated polymer/C, heterojunction photodiodes 3.2. Conjugated polymer/C, tunnel diodes 3.3. Other devices related to photophysics of composite films containing fullerenes 3.4. Single fullerene thin film photodiodes 3.5. Fullerene thin film light emitting diodes: Anomalous electroluminescence 4. Conclusion and Outlook Acknowledgements References

131 131 133 133 137 139 140 140 142 142 143 148 149 I51 153 I53 154 155 157 157

ABBREVIATIONS A BCHA-PPV CB cw

D DC ESR

Acceptor Poly(bisdicholestanoxvparhiphenylene viny&e) Conduction band Continuous wave Donor Direct current Electron spin resonance

FF f-PIA HOMO IT0 LED LESR

131

Fill factor femtosecond photoinduced absorption . Highest occuoied molecular orbital I&urn-Tin-bxide Light emitting diode Light induced electron spin resonance

N. S. Sariciftci

132 LUMO MEH-PPV M-I transition MIM MIS

Lowest unoccupied molecular orbital Poly(2-methoxy,S+?‘-ethylhexoxy)-para-phenylene vinylene) Metal to insulator transition Metal-Insulator-Metal Metal-InsulatorSemiconductor

MO

PC PcM PcZn PIA P30T P3UT TTF uv VB

Molecular orbital Photocurrent Metallophtalocyanine Zinc-phtalocyanine Photoinduced absorption Poly(3-octylthiophene) Poly(3undecylthiophene) Tetrathiafulvalene Ultraviolet Valence band

FOREWORD

At the time, when Editor Peter T. Landsberg requested a “. . . good educational and easy-to-read review article on the potential of fullerenes in a semiconductor context . . .” we completely underestimated the extent of such an undertaking. As a new emerging field, Buckminsterfullerene already shows very promising applications in various types of organic semiconductor devices. A scan of the World Patent Index resulted in 131 fullerene related inventions which were published by the end of April 1994. It is conceivable that during the preparation and publication of this review article this number has already doubled. This a remarkable achievement for such a young field and clearly demonstrates the technological potential. Furthermore, Buckminsterfullerene, C, , is a molecular organic solid, which is a narrow band semiconductor in its pristine form and a metal and even a high temperature superconductor in the heavily n-doped state (reduced by alkali metals such as potassium); for a collection of relevant publications see Ref. (1). The fascinating solid state properties of Cm attracted a large number of scientists to participate in ‘Bucky rush’. Nevertheless, CM is a conjugated molecule and most of its solid state properties can be understood within the framework of z-electron (conjugated) molecular solids. This also explains why a large number of scientists working in fullerene research have backgrounds in different areas of z-electron organic semiconductors. To review all the interesting properties of Buckminsterfullerene and present a general introduction into z-electron organic semiconductors (such as anthracene, phtalocyanine etc.) is beyond the scope of this paper. The interested reader is referred to books on n-electron organic semiconductors.““) In this review we focus on the supramolecular photophysics of two n-electron systems, namely conjugated polymers as donors on one hand and Buckminsterfullerene, C,, as an acceptor on the other. This photophysics, similar to the primary steps in photosynthesis, has been found to be of great interest in the conversion of solar energy into electricity. (‘-“) In this context, we have arranged this article as follows: In the first chapter the conjugated polymeric materials are introduced as n-electron organic semiconductors. Questions about the charge carrier mobility and intermolecular resonance integral are addressed. The very high polarizability of these materials is proposed to enhance the resonance integral in the charged state. Disorder as a major parameter in the physics of these materials is introduced and discussed. The second chapter is a review of the photophysics in the supramolecular composite of conjugated polymers with C,. Specifically, the evidence for an ultrafast, metastable, reversible photoinduced electron transfer from conjugated polymers onto Cm is discussed. This photophysics has been utilized in polymeric semiconductor devices. In the third chapter we briefly review some representative organic, semiconductor devices based on or related to C, in order to demonstrate its potential in future organic solid state technology. We have also included several disclosures from the published patent literature.

Buckminsterfullerenes

in organic photoelectric devices

133

In the last chapter, we discuss the potential for future developments based on interesting device structures such as organic donor-acceptor quantum wells and superconducting microstructures utilizing C, thin film technology.

1. INTRODUCTION

1.1. Conjugated polymeric materials: rc-electron semiconductors

Polymers are typically insulators and generally not considered as interesting from the point of view of electronic materials. This attitude, common to many solid state physicists and electronic engineers, has been revised during the last two decades starting with the discovery of high electrical conductivity of polyacetylene upon doping with iodine vapor.(‘2.‘3) The pZ orbitals of carbon atoms in polyacetylene may hypothetically create a valence band (x-band) which is half-filled due to the fact that every carbon has one pZelectron. This would result in a one-dimensional metal even in the pristine, ground state. However, quasi-one-dimensional metals tend to distort spontaneously such that the spacing between the successive carbon atoms along the chain is modulated with period 2n/kF, where kF is the Fermi wave vector.(‘4) This dimerization of the sp2 hybridized carbon chain opens up an energy gap at the Fermi surface, stabilizing the distortion by lowering the energy; this is the well known ‘Peierfs distortion’ (Fig. 1). Thus, polyacetylene is a semiconductor with an energy gap of around 1.8 eV instead of a one-dimensional metal. Please note, that in Fig. 1, the dimerization is displayed with two different ways of arranging the n-electrons. These two ground state structures are energetically fully degenerate, classifying the polyacetylene as a ‘degenerate ground state’ polymer. This ground state topology creates interesting phenomena such as solitons as domain walls between these two different phases.(‘““) The details of these interesting physical problems go far beyond the scope of this article; the interested reader is

FIG. 1. Schematic description of the Peierls distortion on trans-polyacetylene.

N. S. Sariciftci

134

Polyacetylene (PA)

Polypara hcnylcne (P? P)

Polypyrrole (PP)

Polyparaphenylenc sulphide (PP.9

Polyparaphenylene

Polythiophenc WI

vinylene (PPV

Polycarbazole Poly-3-metflylthiophene (P3MT)

Polyisothianaphene (Pn-N)

tpcB’

-R

“etp^

Poly (I .6- hcptadiync) (PHT)

n

s

R Poly-3 alkylthiophene R- Bury1 --> P3BT R-Ethyl --a P3ET

Polyquinolinc WQ) cdc

s d

Poly-3-alkylsulphonate R’ = CH$ZHzSOjNa (P3ETSNa) R’= (CH,),SO,Na (P3BTSNa)

n

&

s

n

Polyaniline (PANI)

FIG. 2. Molecular structures of the some semiconducting conjugated polymers.

referred to original articles and review papers. u4) Similar to inorganic semiconductors, polyacetylene can be p-doped with electron-accepting impurities (oxidation) or n-doped with electron-donating impurities (reduction). The p-doping of oriented polyacetylene material can lead to electrical conductivities as high as lo4 S/cm along the chain direction.“s-2’) This high value indicated a metallic state for heavily doped polyacetylene. This metallic nature of the heavily doped polyacetylene is further evidenced by the observation of a Pauli paramagnetic susceptibility and Korringa relaxation of the nuclei which results from a finite density of states Recently, several other polyheteroaromatic n-elecat the Fermi energy; i.e. A’(&) # 0. (22-24) tron semiconductors have been synthesized (Fig. 2) which show all the features of the metal-insulator transition in doped inorganic semiconductors yielding a highly conducting metallic state with all the expected magnetic fingerprints for IV&) # O.(25-z*) Thus, it seems to be generally true that polymeric z-electron semiconductors can be viewed in full analogy

Buckminsterfullerenes

in organic photoelectric devices

135

with their inorganic counterparts for the doping induced M-I transition. A comparison of the conductivity regimes of several materials is displayed in Fig. 3. The one-dimensional nature of the polymeric n-electron semiconductors makes them beautiful objects to study quasiparticle excitations such as solitons, polarons and bipolarons. The concepts of the quasi-one-dimensional particles as charge carriers of the conjugated polymeric semiconductors have been heavily studied during the last two decades and the In a simplified scenario the results summarized in various books and review articles. (‘4~2p-3’) M-I transition upon doping proceeds as follows: To start with, the pristine, undoped conjugated polymer exhibits a full valence band (n-band) and an empty conduction band (rr*-band) with the Fermi energy (EF) close to the middle of the gap (Fig. 4a). At low doping levels, the band structure exhibits the characteristic in-the-gap states associated with the doping process. These gap states are located in the 0

[l 1

Heavily p-doped Oriented Polyacetylene

/n.cm]

o6-

copper

lo5 -- Platinum

t

1O4-- Bismuth -- (SN)x 1o3-- Graphite -- -l-E-TCNQ

Heavily doped Polythiophene, Polypyrrolle etc. K-doped C60

102-’

t

10’ -loo-10-l --

Conducting Polymer Blends

I

loS2 -- Germanium lo-3 -lo-4 --

loss

--

Silicon

lo-6 --

A

10-’ -10-a --

Undoped Semiconducting Conjugated Polymers

lo-g --

Undoped C60

lo-lo

7

10-l’ 10

Molecular crystals

-12

FIG. 3. Range of DC electrical conductivities of various x-electron semiconductors compared to some reference materials.

N. S. Sariciftci

136

FIG. 4. Simple band

structural

scheme of the doping

process

in conjugated

polymers

(see text).

middle of the gap for solitons (only in the degenerate ground state polymers such as polyacetylene and its derivatives) and symmetrically split with respect to the center for the polarons and bipolarons (Fig. 4b). The creation for the gap states is associated with a lattice distortion which accompanies the doping process due to the strong electron-phonon interaction in these quasi-one-dimensional rr-electron systems. Upon further doping these isolated quasiparticles start to interact, and the gap states broaden into bands (Fig. 4~). Upon reaching a critical doping level the Fermi energy enters these bands resulting in an insulator-to-metal (M-I) transition (Fig. 4d). Although this simplified picture summarizes the doping process quite well, several important discussions, such as the role of electron-electron correlations, localization, soliton lattice versus polaron lattice, etc. are omitted. The electrical conductivity of heavily doped polyacetylene is as high as 5 x 104S/cm as mentioned earlier. This value is obtained parallel to the chain orientation, which emphasizes the importance of a good exchange or overlap integral along the chain (intramolecular transfer integral). The electrical conductivity perpendicular to the chain direction is at least two orders of magnitude smaller indicating that the overlap between X-orbitals on different chains are indeed much smaller (intermolecular transfer integral). The mobility along and across the chain must therefore be highly anisotropic. Since the macroscopic conductivity is determined by the intermolecular transfer, the temperature dependence as well as the overall value will be dominated by this ‘bottle-neck’ as shown by experiment.@’ Even with this restriction, one can get lo4 S/cm, implying that the intramolecular conductivity parallel to the chain can reach much higher values.(32) Role of disorder: There are, however, limitations to the overall mobility of the charge carriers in conjugated polymeric semiconductors. The fundamental reason for the large concentration of defects in macromolecular solids can be expressed in thermodynamic terms by the relation of the free energy for crystallization per mole of crystallizing unit, AF,, as the sum of two opposing entities; the enthalpy of crystallization, AH,, and the entropy, AS,t33) AF; = AHc - TAS,.

(1)

Compared to the enthalpy of crystallization in covalent, ionic or metallic solids, the enthalpy of crystallization in molecular solids is relatively weak, for it is based only on weak van der Waals intermolecular forces. (33)In addition, highly perfect crystalline polymeric solids have high entropies to pay for e.g. the opportunity of disorder in a long chain with almost freely rotating segments is much higher than a small molecular covalent or ionic crystal. As a consequence of these two unfavorable factors, the low enthalpy of crystallization and the high entropy cost of crystallization, the net free energy of crystallization per crystallization unit is very low in polymeric molecular solids.(33) This tendency toward built-in disorder results in Anderson localization and the associated M-I transitions have been studied deeply in heavily doped polymeric n-electron

Buckminsterfullerenes

in organic photoelectric devices

137

With improvements of material quality semiconductors (conducting polymers). (26*27~34~35) through a high degree of uniformity of chemical coupling, (x+~*)through controlled recrystallization from solutions in the doped conducting state,(3g)and through further refining of the synthesis conditions, semiconducting polymers can now be prepared on the metallic side of This implies that the built-in disorder can be the disorder-induced M-I transition. (26*34*3s) reduced sufficiently to enable it to be used for practical purposes. 1.2. Relation between molecular orbital overlap and transport properties The intermolecular transfer integral is relatively small in the conjugated organic semiconductors, limiting the mobility of the charge carriers as discussed above. How is it then possible that such large values for DC conductivity can be achieved upon doping? Even though that question is not fully answered yet, in this section we will try to illustrate a possible enhancement of the transfer integral due to high polarizability of these a-electron semiconductors. The tight binding approximation considers a collection of interacting atoms with overlap of the wavefunctions and is more easily transferrable to chemical notation of molecular orbital (MO) theories as for example H2+ . For Is, and Is, being the ground state hydrogen orbitals centered on atom A and B, respectively, with the overlap, 01,between these two orbitals a =

(2)

one can write the molecular orbitals and their energies as follows:(*) Y* = [l/(2 + 2a)]“*(lsA + Is,)

(3)

Y = [l/(2 - 2a)]“*(ls* - Is,)

(4)

E* = (P + r)/U + a)

(5)

E=(P -r)/U -4

(6)

where /? =

ls,h,ls,dz..

. the atomic or Coulombic integral

(7)

f

y=

ls,h,ls,dr..

. the resonance integral

(8)

s

and h, the effective Hamiltonian of the system. The resonance (or transfer) integral, y, represents the energy interaction between the orbitals Is, and Is,. The relation between the orbital splitting and the transfer integral, y, may be demonstrated by the energies of the electronic levels for a face-centered cubic lattice constant a: E(k) = Es - p - 12y + #*a*.

A band is formed when k spans all the allowed values. The bandwidth, to the resonance integral, y, by the formula: w = 2&y

(9) W, is thus related

(10)

138

N. S. Sariciftci

with Z, being the coordination number. Mulliken proposed that transfer integral, y, should be roughly proportional to the corresponding overlap integral, a. For the dispersion relation in Sommerfeld theory E(k)

=h2k2/2m*

(11)

with the effective mass m * = h’/(a ‘E (k )/iJk’)

(12)

p = ez,/m*

(13)

and for the mobility

with r, being the carrier scattering time and e the elementary charge, one can formulate the relation of the mobility, p, and the effective mass, m *, to the resonance integral, y, as f0110ws:‘*’ m * = h2/(2ya2)

(14)

p = (2ya2er,)lfi2.

(1%

Thus, a macroscopic experimental parameter, the mobility of the charge carrier, p, is linked to a chemically defined structural feature, the molecular orbital overlap between adjacent molecules. What determines the transfer integral? For estimation of the orbital overlap between adjacent molecules the van der Waals attractive interaction is a good starting point. For non-polar molecules, as is often the case in pristine molecular crystals, the potential profile may be defined by the empirical Buckingham formula(2*40) V(r) = -a/r6 + b exp(-c/r)

(16)

where a, b, c, are constants and r the center-to-center distance between the non-polar molecules. At the van der Waals distance, the potential has a minimum. Due to its short range nature, the attractive interaction falls rapidly with the distance, r. However, for polar molecular systems, there are dipolar and quadripolar terms which can dominate the size of the orbital overlap. For a pair of molecules, A and B, the dipole-dipole interaction energy is given by:” Ed4 = -2/(3kT)p:pk/@

(17)

where pA, pe are the dipole moments of A and B. The dipole-quadrupole interaction energy falls off as l/r’, i.e. is much less important than EdA. Additionally, a non-polar molecule B may interact with a permanent dipole moment 1~~ via the induced dipole? (18) where clg is the polarizability of molecule B. This polarizability and dipole induced transfer integral component can determine the degree of Coulomb screening in conjugated polymers and explain the strong 3-dimensional delocalization of charges in fullerenes and conjugated polymers, even though the ground state overlap between molecules is very small.

Buckminsterfullerenes

in organic photoelectric

devices

139

Of course, the above discussion is far from being complete. The Coulomb correlations of opposite charges, electron-electron interactions, as well as the theory of localization and effects of disorder must be considered. We will readily use the simple notation as discussed above, to develop simple guidelines for molecular semiconductors.

1.3. Why n-electrons? In sp2 hybridized chemical structures the spatial orthogonality of the pZ electrons to the other sp2 hybridized sigma (a) orbitals is important for the one-electron picture described above. The idealized n-electron structures of polyenes are demonstrated in Fig. 1. The resonance structures drawn in chemistry books for benzene rings as well as the two dimensional analogue of these strongly coupled n-electron structures in graphite can be used to visualize the importance of the rr-electrons in the field of organic semiconductors. Actually, the carbon-to-carbon distance in a layer of graphite is only 1.41 A, intermediate between a single (1.54 A) and a double (1.33 A) bond, giving rise to a strong overlap between the p, electrons and completely delocalizing them. The strong overlap also results in relatively large mobilities, such as ,u x 13,000 cm2/Vs at room temperature in graphite.c2) An important aspect of rc-electron system with large degree of delocalization is the high polarizability of the electron cloud. Hiickel-type MO calculations for a linear regular (no bond alteration) polyene chain result in infinite longitudinal polarizability per unit in the limit of an infinite chain length. c3’)This theoretical prediction is a simplified one. Several other factors limit the increase of the polarizability with increasing n-electron chain length, such as the bond alternation (Peierls distortion), disorder etc. (‘I)Nevertheless, extended x-electron systems such as polyenes and carbocyanines exhibit relatively large polarizabilities compared to organic molecules with saturated bonds.“‘) This high polarizability is important in the context of the transfer integral related to the dipole-induced dipole interaction in a charged rc-electron system (Eqn. (18)). In analogy with inorganic, direct gap semiconductors, it is also convenient to look at the energy gaps of the n-electron materials. The bonding and antibonding rr-orbitals of an sp2 hybridized n-electron material (such as polyenes) create energy bands which are fully occupied (n-band) and empty (n*-band) in full analogy to the valence band (VB) and the conduction band (CB) in inorganic semiconductors. (4’) The energy gaps in polyenes, polyaromatics and several n-electron materials cover the entire visible regime of l-4 eV. This is distinguished from organic systems with saturated bonds where the bonding and antibonding orbitals give rise to bands with energy gaps far beyond the visible regime (Eg > S eV), classifying these systems as insulators rather than semiconductors. At first glance, the physics of pristine fullerene crystals is similar to other n-electron materials with classical examples such as napthalene, anthracene, phtalocyanine, charge transfer salts etc.c2”@) Some important differences, however, should be pointed out: The overlap between the individual small molecules within a n-electron molecular crystal (such as anthracene) is weak; whereas in an infinite polyene, the intramolecular n-electron overlap is fairly high, resulting in delocalization of wavefunctions along the molecule as well as high longitudinal polarizability. As the molecular units within a n-electron molecular crystal get larger and larger, a high conjugation polymeric crystal on one hand and graphite on the other hand will result in one and in two dimensional limits, respectively. Undoped, solid fullerenes, however, are n-electron molecular solids consisting of a very large number of n-electron units which are ‘bent in space’ to result in cyclic boundary conditions. Thus, they will carry some physics of the one and/or two dimensional ~-electron semiconductors in addition to the molecular crystalline properties. In this context, fullerenes combine the advantages of a large n-electron system with all the interesting physics; and they

N. S. Sariciftci

140

get rid of the built-in disorder of conjugated polymers (as discussed above) by forming defined crystalline structures.

2.

PHOTOINDUCED POLYMERS

ELECTRON TRANSFER FROM ONTO BUCKMINSTERFULLERENE,

CONJUGATED C,,

If one mixes Buckminsterfullerene, C,, into conjugated, semiconducting polymers, there are very interesting photophysical phenomena in these supramolecular composites. We reported the discovery of a photoinduced electron transfer from semiconducting conjugated polymers onto C,, which is reversible, ultrafast (electron transfer occurs within 300 fs!), and metastable (charge separated state can be stabilized at low temperatures). Figure 5 schematically visualizes the observed phenomenon. The photoinduced electron transfer from conjugated polymers onto C, is one of the fastest ever reported (the quantum efficiency of this process is close to unity!) and therefore exhibits a very robust and competitive route for applications which utilize photoinduced charge separation. 2.1. Background The parallel interests of understanding natural photosynthesis in biosystems and of efficiently harvesting solar energy as an alternative to fossil fuels have led to a substantial, multidisciplinary effort in the field of photoinduced electron transfer phenomena in physics, chemistry and biology as well as their overlapping regimes.(42A3’A basic description of intermolecular and/or intermolecular photoinduced electron transfer is as follows:

Step Step Step Step Step

1: 2: 3: 4: 5:

‘v3D*+ A, (excitation on D); D+A+ ‘s3D*+ A + ‘J(D-A)*, (excitation delocalized on D-A complex); ‘*3(D-A)* + ‘*3(D6+-A6-)*, (charge transfer initiated); ‘.3(Dd+_Ab- )* + ‘*3(D+‘-A-‘), (ion radical pair formed); ‘.3(D+‘_A-‘) ---, D+’ + A - ‘, (charge separation);

FIG. 5. Schematic illustration of the photoinduced electron transfer from semiconducting polymers onto c,.

Buckminsterfullerenes

in organic photoelectric devices

141

where the donor (D) and acceptor (A) units are either covalently bound (intramolecular), or spatially close but not covalently bonded (intermolecular); 1 and 3 denote singlet or triplet excited state, respectively. The partial charge transfer at Step 3 is strongly dependent on the surrounding medium, such as the polarity of the solvent etc. resulting in a continuous range for the transfer rate 0 < 6 < 1. At Step 4, a whole electron is transferred, i.e. 6 = 1. At each step, the D-A system can relax back to the ground state either by releasing energy to the ‘lattice’ (in the form of heat) or through light emission (provided the radiative transition is allowed). Permanent changes that may occur from ion radical reactions beyond Step 5 are not considered here, even though their importance in photochemical and photoelectrochemical reactions has been established. The electron transfer (Step 4) describes the formation of an ion radical pair; this does not occur unless Z E - A, - UC < 0, where Zz is the ionization potential of the excited state (D*) of the donor, AA is the electron affinity of the acceptor and Uc is the Coulomb energy of the separated radicals (including polarization effects). Stabilization of the charge separation (Step 5) can be enabled by carrier delocalization on the D+ (and/or A-) species and by structural relaxation. A highly polar environment can also stabilize charge separation by screening the separated charges. The possibility of using such charge separation in molecular information storage and optoelectronics has been suggested. (44*45) Donor-bridge-acceptor type ‘supermolecules’ have been proposed as bistable ‘molecular information storage units’, in which the separated ion radical pair state is visualized as one logic state and the ground state is the second logic state.(46)For intramolecular photoinduced electron transfer, the requirements on the spacer between the donor and acceptor units are important (and demanding); for metastability, the molecular orbitals of the D and A components must be decoupled so as to retard the back electron transfer process.(47.48) Light harvesting polymer systems utilizing photoinduced energy and/or electron transfer mechanisms attract more and more attention in the scientific community active in this field.(49) Polymers are particularly interesting in this field of photophysics and photochemistry because new synthetic methods make it possible to modify, functionalize and derivatize donor and acceptor units within the framework of polymer chemistry and materials science. Light sensitive polymer arrays are considered for photocatalysis, non-silver based imaging, all optical information storage, electroresponsive systems for displays and sensors among many others (42%X’Ul) There have been several reports on photophysical phenomena related to photoinduced electron transfer between conjugated polymers and C,. Wang reported enhanced photoconductivity in polyvinylcarbazole(52) and polysilane(53) mixed with C,,. The mixing of Buckminsterfullerene with donors such as dialkylanilines result in ultrafast photoinduced charge transfer phenomena as reported by Hochstrasser and his group.‘54’ These authors report the quenching of the intersystem crossing due to ultrafast charge transfer from dimethylaniline onto C&. In their experiments, excitation is performed on the Buckminsterfullerene (acceptor) across the forbidden energy gap resulting in hole transfer which is analogous to the donor excited electron transfer as described in Steps l-5 above. Wang and Cheng reported the formation of a charge transfer complex in the excited state”” (exciplex, 6 < 1), rather than a whole photoinduced electron transfer (6 = 1). By comparison of toluene with dialkylaniline, a strong increase in non-linear optical response is observed for the latter due to its strong donor properties.‘56) Kepler et al. also reported photoinduced electron transfer from polysilanes onto C,.‘57) Another example of photoinduced charge transfer phenomena with the participation of C, is presented by Prashant V. Kamat by utilizing ZnO suspension composites with Buckminsterfullerene.‘58’ Furthermore, recent reports on photoinduced polymerization of C,‘59) and photoinduced electron transport across lipid bilayers mediated by C,(60’ stress the importance of Buckminsterfullerene in photophysical and photochemical phenomena.

142

N. S. Sariciftci

K. Yoshino, A. A. Zakhidov and coworkers reported studies on mixtures of poly(Zalkylthiophene)/C,@‘). They observed a strong quenching of semiconducting polymer luminescence and interpreted the results as due to doping effect of C, in the ground state. Doping of semiconducting polymers by C, is ruled out in our studies due to absence of CU anion electron spin resonance signal in the ground state. The same group also reported the enhancement of near steady state photoconductivity and quenching of luminescence of several conjugated polymers mixed with C, supporting the ultrafast, reversible photoinduced electron transfer between semiconducting polymers and Cm .(6147) 2.2. Direct experimental evidence for the metastable charge separation: Light Induced Electron Spin Resonance (LESR) Definitive evidence of charge transfer and charge separation is obtained from LESR experiments.“) Figure 6 shows the ESR signal upon illuminating the conjugated polymer/C, composites with light of hv = E ,*-, where E & is the energy gap of the conjugated polymer donor component. Two photoinduced ESR signals can be resolved; one at g = 2.00 and the other at g = 1.99.(‘JEWThe higher g-value line is assigned to the conjugated polymer cation (polaron) and the lower g-value line to C, anion. The assignment of the lower g-value line to C& is unambiguous, for this low g-value was measured earlier for C&@; the higher g-value is typical of conjugated polymers. The LESR signal vanishes above 200 K; this rules out permanent photochemical changes as the origin of the ESR signal and demonstrates the reversibility of the photoinduced electron transfer. The temperature dependence of the LESR signal intensity shows and Arrhenius behavior with activation energy of AE x 15 meV. This result suggests a phonon assisted back relaxation mechanism of the photoinduced charge separated state. 2.3. Direct observation of 300 femtosecond forward electron transfer(‘0*70) Although the time scale for the forward electron transfer can be estimated through the demonstration that the luminescence is quenched by reduction of the associated luminescence decay time, the effect of non-radiative decay channels other than the electron transfer cannot

Magnetic

Field [G]

FIG. 6. LESR spectra of P30T/C, upon succesive illumination with 2.41 eV Argon ion laser with IOOmW. The dark ESR signal is drawn as a dashed line and the LESR signal as a solid line.

Buckminsterfullerenes

in organic photoelectric devices

143

be ruled out. Thus, it is important to directly observe the electron transfer by femtosecond photoinduced absorption (f-PIA) spectroscopy. The f-PIA spectrum at several delay times for pure poly (3-octylthiophene) (P30T) is shown in Fig. 7a. The rise time of the absorption is resolution limited (< 300 fs) and reaches maximum at 750 fs. The spectrum has two distinct peaks, one at 1.9 eV and the other at 1.3 eV (the oscillations are an artifact due to spectral oscillations in the continuum). The initial decay time is approximately 1.2 ps. However, even at 750 fs the P30T/C,, composite f-PIA spectrum (Fig. 7b) exhibits a broad absorption centered around 1.6 eV which is different from Fig 7a. This broad maximum subsequently shifts from 1.6 eV toward the red to 1.4 eV (after 500 ps). The decay curve of the 1.6 eV absorption in the P30T/C, composite is much different from the decay of 1.9 eV absorption in the pure P30T material. A metastable state is reached after 2-3 ps following photoexcitation (decreasing by only a factor of 3 after 500 ps). Thus, we conclude that the electron transfer takes place at times less than 1 ps.

To actually resolve this forward electron transfer time scale the photoinduced dichroism has been utilized.“‘) The dichroic ratio is defined as the ratio of the photoinduced absorption with the pump and probe polarization parallel to that with the polarization perpendicular. For electron-hole pairs initially excited on single polymer chains the dichroic ratio should be 3.(“) As the excitation diffuses to other molecules, this polarization memory will be lost resulting in a dichroic ratio of unity. Thus, upon photoinduced electron transfer the polarization memory time will be quenched with the electron transfer rate. As shown in Fig. 8, adding 1% of C@ into conjugated polymers results in the decrease of the dichroic ratio from 3 to 1.5 within 300 fs compared to the tens of picoseconds in the pristine material. This result shows the electron transfer time to be approximately 300 fs. To demonstrate that ultrafast electron transfer yields a metastable charge separation which was observed in near steady state PIA experiments and LESR studies, we compare in Fig. 9 the photoinduced absorption spectra of P30T/Ca composites in two different time domains, one with millisecond chopped excitation and the other with femtosecond resolution. It is clearly observable that the charge separated state is created at very early times and lives milliseconds at 80 K by showing the same excited state absorption spectrum in 5 ps as in milliseconds. 2.4. Sensitization of photoconductivity Figure 10 shows the time-resolved transient photocurrent (PC) of a P30T film containing 5% &, and a pure P30T film upon photoexcitation at 2.9 eV. The addition of 5% C, to P30T results in an enhancement of the photocurrent by nearly an order of magnitude. The rise time is limited by the temporal resolution of the instrumental response (50 ps). The increase of the photocarrier generation efficiency results in the enhancement of the initial photocurrent. The generality of this enhancement for other non-degenerate semiconducting polymers is demonstrated in Fig. 11 for the case of MEH-PPV/C,O composites with different C,,, content. The admixture of 1% of C, results in an increase of initial photocurrent by nearly an order of magnitude. This increase of the photocarrier generation efficiency is accompanied by an increase in lifetime of the photocarriers upon adding more C,,. Thus, the ultrafast photoinduced electron transfer from the semiconducting polymer onto Cm not only enhances the carrier generation, but also serves to prevent recombination by separating the charges and by stabilizing them. Figure 12 shows, on a semilog plot, the spectral response of the cw-PC of MEH-PPV alone and of MEH-ppv/C, composites for different concentrations of C,. These room temperature data are normalized to constant incident photon flux of about 7.5 x lOI photons/cm* s. The photoconductive spectral response of MEH-PPV shows a sharp onset at about 2 eV which

144

N. S. Sariciftci

‘\ . C.

A

1.4

1.6

1.8

I

LL

2.c

Energy [eV] (4

FIG+ 7. (a) Femtosecond photoinduced absorption spectrum of P30T at various delay times after excitation with a 100 fs, 2.01 eV pump pulse. The inset shows the electronic structure of the neutral bipolaron (polaron+xciton). (b) Femtosecond photoinduced absorption spectrum of P30T/C, (50%) at various delay times after excitation with a 100 fs, 2.01 eV pump pulse.

Buckminsterfullerenes

0.1

in organic photoelectric devices

10

1

145

100

Time [ps] FIG. 8. Room temperature dichroic ratio of BCHA-PPV at 1.41 eV (open circles) and of BCHA-PPV/C, (1%) (closed triangles). The open diamonds display the dichroic ratio of MEH-PPV at 1.5 eV and the closed squares for MEH-PPV/C, (I%), also at 300 K.

coincides with the optical absorption edge across the energy gap. However, the photoconductive response of the MEH-PPV/C, composites increases sharply at about 1.3 eV which is lower than the photoconductivity onset of the components alone. Furthermore, the composite films exhibit a remarkably enhanced photoconductivity over the entire spectral region from near infrared to ultraviolet. Especially, at photon energies below 2 eV, the enhancement reaches several orders of magnitude. Although the effect of the C, on the mobility of the carriers should be considered (we suggest that trapping by C, will reduce rather than enhance the mobility), the significant effect of Cso on the charge carrier generation efficiency can be clearly recognized by the observation that even incorporation of 1% C, into the polymer matrix enhances the cw-PC by more than an order of magnitude. This observation is in full agreement with the photoinduced electron transfer phenomenon which leaves metastable positive polarons on the polymer backbone after the electron transfer; e.g. quasi photo-

0.8

Energy

(eV)

FIG. 9. Comparison of the photoinduced absorption spectra for near steady state (millisecond) and ultrafast (picosecond, Fig. 7) time domains for P30T/C,.

146

N. S. Sariciftci

0

300

200

100

Time FIG. 10. Transient photoconductivity

(ps)

in P30T alone (full circles) and in P30T sensitized with 5% C, (open circles).

A 0 0 0

lo2

5o%c60 58C60 18C60

haH-PPV

I

1

10’

loo

10-l

Time (ns) FIG. 11. The transient photoconductivity of MEH-PPV/C, mixed in MEH-PPV.

for various concentrations

of C,

Buckminsterfullerenes

in organic photoelectric devices

147

Energy (eV) FIG. 12. Spectral response of the steady state photoconductivity of MEH-PPV alone and MEH-PPV/C, for several concentrations of C,. The data were taken at room temperature under a biasing field of 104V/cm.

induced doping. The scenario of photoinduced electron transfer using a band structure scheme is displayed in Fig. 13. This molecular-based photoinduced charge transfer has been utilized in semiconductor polymer Cso heterojunctions to fabricate photodiodes and photovoltaic devices.@) We

Vacuum

4- + I

29

ev

Lumo

VB

Homo

+t

Conducting polymer

C60

FIG. 13. Schematic energy band diagram for the photoinduced electron transfer from semiconducting polymers onto C@. (See list of abbreviations.)

N. S. Sariciftci

148

summarize the characterization of rectifying heterojunctions (diodes) fabricated from the semiconducting polymer (MEH-PPV) and C, in Chapter 3 below. 3. BUCKMINSTERFULLERENE,

CsO, IN ELECTRONIC

DEVICES

The devices for photovoltaic energy conversion mostly consist of semiconductor/metal interfaces (Schottky junction) or p-type/n-type semiconductor junctions (heterojunctions). Classical Schottky junction: When a metal and an n-type (for example) semiconductor are brought together, electrons flow from the semiconductor to the metal until the Fermi levels are equalized. The space-charge region is the electron depleted regime at the semiconductor side of the interface and is made up of remaining positively charged ionized impurities. The characteristic parameter of the Schottky junction are: Qr,,,= E,,, - Er(M) work function of the metal

(19)

@s = E,,, - EF(S) work function of the semiconductor

(20)

with EVBcand EF being the vacuum energy level and the Fermi energy, respectively. The current-voltage characteristic for Schottky junctions obey a Shockley type equation with thermionic emission.(2~72) J = J,,,[exp(eva/kT)

- l]

(21)

Jsat= A* * T* . exp(AEM,/kT)

(22)

AEMs = @, - As

(23)

A, = Ew - E,

(24)

where: I’,: applied voltage; J: current; Jsa,: saturation current; A*: Richardson constant; AEMs : energy barrier; A,: electron affinity of the semiconductor; E,: conduction band edge. Classicalp-n heterojunction : The intrinsic charge carrier concentrations of a semiconductor arise from thermal excitation of the charge carriers across the energy gap. Upon doping, new energetic levels are introduced in the gap and the Fermi energy is no longer in the middle of the gap, but displaced towards the conduction band (n-doping) or the valence band (p-doping). If two parts of a semiconductor, one p-doped and the other n-doped, are joined at an interface, electrons flow from the n-type to the p-type semiconductor creating an electric field V,,i (built-in potential) at the interface. The area of ionized impurities is called the space-charge or depletion region. The Fermi energies on both sides must be equal at equilibrium. The energy shift needed to equalize the two Fermi levels is: e. l’bi = EF(n) - EF(p).

(25)

Are Conjugated Polymer Devices Dlfirent? Molecular crystals or conjugated polymer systems have very low intrinsic conductivities. For example, with an energy gap of 4 eV, which is necessary for a polymeric blue light emitting diode, the fraction of electrons excited across the gap at room temperature is of the order of 10e3’! This has dramatic consequences for the description of the metal-polymeric semiconductor junctions as well as for the organic heterojunction interface: (1) At the metal/polymer junction ((I$, > A,), there is negligible charge exchange and the extent of space charge region and the band bending is very small.

Buckminsterfullerenes

in organic photoelectric devices

149

(2) At the p-n heterojunction,

there are no charge carriers available to flow across the interface at room temperature; thus the depletion layer is confined to very few monolayers at the interface with nearly vanishing built-in potential.

In other words, the metal/polymer junctions can be described as metal/insulator (MIM) rather than metal/semiconductor junctions. 3.1. Conjugated polymer/C,

heterojunction

junctions

photodiodes

Considering the energy band diagram in Fig. 13, it is immediately clear that the heterojunction formed at the interface between a semiconducting polymer and C, thin film should exhibit rectifying current-voltage characteristic (e.g. analogous to a p-n junction). One polarity of such a device (where electron injection on the semiconducting polymer side and electron removal on C,) is energetically unfavorable. This polarity results in very low current densities. On the other hand electron injection onto Cm and electron removal from semiconducting polymer is energetically favorable, creating conducting species on both sides. This polarity of the device results in relatively high current densities. Similar consideration of ‘molecular diodes’ were first proposed by Aviram and Ratner years ago for Langmuir-Blodgett D-A structures.(44) Figure 14 shows the current voltage characteristics of a heterojunction device consisting of ITO/MEH-PPV/C,/Au; the schematic description of the device@~“’is shown in Fig. 15. Positive bias is defined as positive voltage applied to the IT0 contact. Exponential turn-on up to 500 mV in forward bias is clearly observable; the rectification ratio is approximately 104. In order to test if the IT0 or the gold electrodes form blocking (or rectifying) contacts, the following three layer devices were prepared: gold/MEH-PPV/gold, gold/MEH-PPV/ITO, gold/C,/gold, and gold/C, /ITO. Since all such devices have completely symmetric and linear current-voltage characteristics, we conclude that gold and IT0 form non-rectifying contacts both to MEH-PPV and to Cm, in agreement with the reported absence of rectification in Au/C, /ITO devices. (73)We therefore, attribute the rectifying behavior of the four layer device to the interface of the hetkojunction between the semiconducting polymer and C,. The current-voltage characteristic of the device changes dramatically upon illumination with visible light. Figure 16 shows the current-voltage data, with the ITO/MEH-PPV/C,/Au device in the dark and with the device illuminated with 514.5 nm light with intensity (Pi,) of x 1 mW/cm*. The open circuit voltage (I’,) is 0.44 V which saturates to around 0.53 V under

Probe to IT0

-2

-1

0

1

Probe to Front Electrode (Gold)

2

BiasM FIG. 14. Dark current vs. voltage characteristics of the ITO/MEH-PPV/C,/Au device at room temperature.

Conducting

Polymer

FIG. 15. Schematic cross-section of the heterojunction devices fabricated from MEH-PPV and C,.

N. S. Sariciftci

150

stronger illumination. The short circuit current density (J,) is 2.08 x lop6 A/cm’. The fill factor (FF) can be obtained from the relation

FFcJO JscVoc

(26)

’

The data in Fig. 16 gives a fill factor of 0.48 and a power conversion efficiency of 0.04%. An open circuit voltage of about 0.5 V appears to be characteristic of the MEH-PPV/C, interface. Similar values were obtained with ITO/MEH-PPV/C,/Al devices and with Al/MEH-PPV/C,/ITO devices. As expected from the observation of fast photoinduced electron transfer from MEH-PPV to C6o,(7,74) a major increase in both forward and reverse bias current is observed to result from photoinduced charge separation at the heterojunction interface. The jump in the current by nearly four orders of magnitude (from 1 x lo-’ A/cm* to 6 x lop6 A/cm’) upon illumination at - 1 V (reverse) bias demonstrates that the heterojunction serves as a relatively sensitive photodiode. Figure 17 shows the dependence of the short circuit current and the photocurrent at - 1 V (reverse) bias as a function of the illumination intensity (514.5 nm). The data in Fig. 17 show no indication of saturation at light intensities up to approximately 1 W/cm*; i.e. one order of magnitude greater than the terrestrial solar intensity. The spectral dependence of the photoresponse of these heterojunction devices is displayed in Fig. 18. The onset of photocurrent at hv z 1.7 eV follows the absorption of MEH-PPV, which initiated the photoinduced electron transfer;(7,74)note that illumination is from the ITO/MEH-PPV side of the device. The minimum in the photocurrent at hv w 2.5 eV corresponds to the energy of maximum absorption of MEH-PPV. We propose, therefore, that the MEH-PPV layer acts as a filter which reduces the number of photons reaching the MEH-PPV/C, interface. This implies that diffusion of charge carriers in these devices is limited; the photoactive region is restricted to a thin layer adjacent to the interface between the MEH-PPV and C, layers. Assuming a charge carrier mobility of z 1O-4cm*/Vs, with an applied field of lO’V/cm and with a photoinduced carrier lifetime z 1 ns, the maximum distance a carrier could travel in the MEH-PPV is estimated to be only a few Angstroms. I

2

I

I

I

I

I

~_ 1o-3 5 3 .g 1om5 O0

E

ooo”

2 E g 1o-7 3 -8

-0

10” -1

0

1

Bias[v] FIG. 16. Current vs. voltage characteristics of the ITO/MEH-PPV/C,/Au device in the dark (diamonds) and upon illumination with the 514.5 nm line from an Argon ion laser of 2 1 mW/cm* (triangles).

I 1oe5

00

..

0.

2. I

I 1oe3

I

10"

Light Power [W/cm’] FIG. 17. Short circuit current (closed circles) and photo current at - 1 V bias (open circles) as a function of light intensity for the ITO/MEH-PPV/C,/Au device.

Buckminsterfullerenes 1.6

I

1.4 s k 3 g

5’ 2

Q

I

1.0 -

cc 00 . . 0.

0.8 -

:

1.2 -

06.

!

0.4 -

‘. +. . s ..a;

in organic photoelectric devices

’ . ..

.

.

-

l 2

;

Wire to IT0

1.5

Wire to Front Electrode

b

f

0.2 -

0.0 .

151

I 2

I

I

I

2.5

3

3.5

Metal 4

Glass

Energy WI

(with/without C60)

FIG. 18. Spectral response of the photocurrent in ITO/MEH-PPV/C,/Au photodiode at (reverse) - 1 V bias.

FIG. 19. Schematic cross-section of the tunnel diodes fabricated from MEH-PPV with C,.

Thus, the generation of photoexcitations which result in separated charge carriers occurs at the heterojunction interface. Yoshino and co-workers also reported the optical response of a heterojunction device comprising a P30T and C, bilayer. (W The photoresponse of these devices shows a broad excitation profile ranging from 750nm into the UV. For quantitative comparison of the rectification ratios and short circuit current, device parameters in the described P30T/& heterojunction devices are needed. Yamashita and coworkers reported a bilayer photodiode based on the organic donor tetratiafulvalene (TTF) and C, .(“) The dark characteristic of devices of the type Au/TTF/C,/Au do not exhibit rectification. Upon illumination, however, the I-V characteristic changes dramatically, exhibiting a large rectification ratio as well as photocurrent.

3.2. Conjugated polymer/C,

tunnel diodes

For a junction between a low workfunction metal and a conjugated polymer we will use the notation ‘tunnel diodes’ due to a low charge carrier concentration with negligible bend bending and barrier formation in those devices as described by Parker.“@ Conjugated polymer tunnel diodes have been studied extensively due to their electroluminescent light emitting (LED) properties which have attracted much attention over the last four years.‘7”89)Schematic cross-sectional view of these devices are displayed in Fig. 19. In forward bias these devices exhibit relatively high efficiency electroluminescence; they are promising for flat panel and/or flexible, large area low cost display applications. In reverse bias, on the other hand, the devices (for example using MEH-PPV) exhibit a strong photoresponse with a quantum yield > 20% (el/ph at - 10 V reverse bias). ca9)Tunnel diodes based on derivatives of polythiophene exhibit even better photoresponse (80% el/ph at - 15 V), competitive to UV sensitized Si photodiodes.(W) There is also a photovoltaic response observed at zero bias.‘89,9’JThe energy band structure for such thin film conjugated polymer devices can be approximated by the rigid band structure model as displayed in Fig. 20. (‘@Since the intrinsic charge carrier concentration in these undoped, pristine materials with an energy gap of > 2 eV is extremely low, there is negligible band bending at the metal-polymer interface. The devices with a typical thickness of 0.1 pm can be described by tunneling of charge carriers at the metal-polymer interface.‘76’ However, it should be considered that the mobility of injected charge carriers is not symmetric, e.g. in most conjugated polymers the hole transportation is favored and thus, electron mobility is lower than the hole mobility.

152

N. S. Sariciftci

a

FIG. 20. Flat band scheme of the Ca/polymer + C,/ITO tunnel diodes at open circuit and short circuit conditions.

Adding C& into the conjugated polymer matrix in Ca/MEH-PPV/ITO devices has a profound effect on the photovoltaic performance of the devices.(“) The mixing of 5% (weight) C& into the MEH-PPV results in enhancement of the short circuit current (1,) by nearly two orders of magnitude.(92) The corresponding quantum yield (el/ph) is as high as 16% compared to 0.09% for the pure MEH-PPV device. c9’) The physical origin for this remarkable enhancement is the above described photoinduced electron transfer effect which occurs at a sub-picosecond time scale inhibiting the geminate or early time radiative and non-radiative recombination processes. The result is a profound enhancement of the quantum yield of the photogenerated charges with an accompanied enhancement of the photoinduced carrier lifetime due to the deep trap effect of the Cm within the conjugated polymer matrix.(93’ Another important effect of adding C, into the MEH-PPV matrix in tunnel diodes is the reduction of the open circuit potential (V,) which is 1.6 V and 0.8 V for the pure MEH-PPV and MEH-PPV + C, layers, respectively. (92)This can be understood as resulting from pinning of the Fermi energy at the lowest occupied molecular orbital (LUMO) of the C, which acts as an impurity level within the gap, whereas in pure devices the V, is roughly the work function difference between the two contacts (Ca and ITO) indicating no intrinsic defect states at the surface of the pure polymer film. (92)The dark characteristic of the MEH-PPV + Cm tunnel diode is similar to the pure polymer devices indicating no ground state interaction between conjugated polymer and C, in full agreement with the above described results W’V8,7WW’4) Tripathy and co-workers reported the enhancement of the photocurrent upon mixing 10% C& into poly(3undecylthiophene)(P3UT) in AI/P3UT + C&TO devices,“‘) but did not quantify the device parameters, thus we are unable to compare device performances. In their studies there is considerable degradation of the device characteristic upon adding 30% Cm into the polymer which can be explained by the morphology effects of a phase segregated and percolated Cm component in the polymer host. The low solubility of pure C, causes a phase segregation during the film preparation process of the conjugated polymer/C, composites. High concentration composites (like 1: 1 weight ratio) almost always result in percolated phase segregation and the device quality is greatly reduced. Shorted devices arising from non-uniformity of the composite films are commonly encountered problems. Thus, there is a trade-off between the advantages of maximizing the C, loading and film quality requirements. An acceptable compromise seems to be composite films of reproducible quality with Cm contents up to 5% (weight) Cm in polymer matrix. Another approach towards eliminating this problem is synthetic chemical funtionalization of the Buckminsterfullerene molecule. This approach (as realized by the group of Professor F. Wudl at the University of California, Santa Barbara) results in enhanced solubility of the Bucky derivatives in common organic solvents and allows higher loading capacity into polymer host matrices.

Buckminsterfullerenes

in organic photoelectric

devices

153

3.3. Other devices related to photophysics of composite films containing fullerenes In this section, we want to briefly review a selection of other patented applications related to the photoinduced electron transfer using Cm. The enhancement of the photoinduced charges has been utilized by the patent of the Xerox Corporation U.S.A. (inventor R. F. Ziolo) in the toner for electrostatic imaging in the xerographic process using resin particles such as styrene methacrylate, styrene butadiene or styrene acrylate and pigment particles such as carbon black, magnetites and photoenhancing additive comprising fullerenes. (96’Images of good resolution are reported. Furthermore, a photoconductive imaging member, comprised of a supporting substrate (can be metallic), a photogenerator layer and a charge transport layer, where fullerenes can be in the charge transport layer as well as in the photogenerator layer, has been patented by of Cm yields sufficient the Xerox Corporation. ‘97) In this application, photoexcitation photogenerated charge carriers for the photodischarge of an imaging member. In a separate report, Mort et al. give the photoinduced charge generation efficiency as 5 x lop3 in the visible (725-550 nm) and 6 x lo-* at 350 nm roughly proportional to the absorption coefficient.(98’ Another approach for photolithographic purposes has been patented to Nippon oil Co., Japan (inventor Noubo Aoki), utilizing a product of the reaction between Cm and n -propylamine of methacryloyl.(99’ The enhanced non-linear optical properties of charge transfer complexes of fullerenes with N,N-diethylaniline is claimed in the patent by DuPont de Nemours and Co., U.S.A. (inventor Ying Wang), which covers the charge transfer complexes and non-linear optical devices based on them.(‘O@Furthermore, DuPont patented (inventor Ying Wang) charge transfer complexes comprising fullerenes and electron donating components and photoconducting composites containing fullerenes. (lo’) Several photoconductive polymers including polyvinylcarbazoles, polysilanes etc. are mixed with fullerenes and the cast films have been used for Corona charge/light induced discharge experiments for photolithographic and xerographic purposes. The enhanced charge generation is again due to electron transfer processes between these donors and fullerenes as acceptors. Again, photoreceptors containing fullerene clusters for electrophotographic purposes is the topic of the patent by Idemitsu Kosan Co., Japan (inventor Shigematsu, Kazuyoshi).(“*’ A photoreceptor for high speed copying, containing a pigment-dispersed photosensitive layer containing a charge generating substance, a hole transporting substance and a carbon cluster with a closed shell structure (such as Cm and C,) has been presented. It is clearly observable, that several applications, ranging from photovoltaic devices up to xerographic photoreceptors, have been realized utilizing the strong photoacceptor properties of fullerenes. This photophysical approach can be viewed as potentially promising for technological applications of fullerenes. 3.4. Single fullerene thin film photodiodes Such structures are similar to the diodes described above and consist of a single layer of fullerene thin film sandwiched between metal electrodes. Due to the small bandgap, the intrinsic charge carrier generation is similar to silicon. In their patent publication, Hebard and co-workers at the AT&T Bell Labs,(‘O”describe the photodiode based on a thin film of Cso which exhibits photoconductivity in a Ag/C,,/Ag surface configuration. Due to identical metal contacts there was no photovoltaic effect observed. However, upon slight doping with potassium (n-doping) the thin film exhibits photovoltaic properties due to Schottky barrier formation by injected charge carriers at the metal interface. Furthermore, the immersion of solvent cast thin films of Cm into electrochemical cells results in photovoltaic effect upon illumination of the electrode, typical of n-type semiconductors in liquid junction cells, as reported by the same group.(‘03.‘04’

1.54

N. S. Sariciftci

Yonehara and Pat reported the photovoltaic effect on a Al/C60/Au type sandwich device.(73) The devices exhibit quantum efficiencies for photoconductivity as high as 53% which is several orders of magnitude higher than earlier reports on Cm sandwich cells.(“‘) The observed excitation profile of the photocurrent closely follows the optical absorption. The photocurrent decreases upon increasing the incident light intensity above a threshold value of 3 x lo-’ W/cm2. Yonehara and Pat also describe an interesting oxygen effect on the characteristic of the devices. The devices without any exposure to oxygen in all the preparation steps exhibit no rectification. The devices exposed to air exhibit much better rectification ratios (66 at -2 V) in the dark current-voltage characteristic as well as greatly enhanced photoresponse. The authors interpret this effect due to the formation of an oxide layer at the Al metal surface resulting in metal-insulator-semiconductor (MIS) structures rather than Schottky devices. This effect of oxygen is not new to the field of molecular semiconductors. For example, metallophtalocyanine (PcM) thin film devices in sandwich configuration (Al/PcM/Au) do not exhibit any rectification if well protected from oxygen, despite the fact that the two metal electrodes have large difference in workfunctions. (2)Upon exposure to air, rectification ratios up to 10’ have been observed. Oxygen exposure also results in rectification ratios up to lo4 in Au/PcZn/Au devices where symmetrical, non-oxidable metal electrodes are used.(‘) As a result of systematic investigations, the effect of an oxide layer on the metal electrode has been found to be negligible, discarding the argument of MIS structure in those cases. The effect is rather on the charge generation efficiency in the PcM film upon transferring one electron onto oxygen molecule (e.g. O2 photodoping). c2)Whether this is similar to the Cm devices of Yonehara and Pat remains to be settled. Curran et al. reported the formation of Schottky barrier and the rectifying current-voltage characteristic for the Au/C,,/n-Si, ITO/C,/Ca/Al and ITO/C,/Mg-In devices.(‘06) For the case of the n-Si Schottky device a small photocurrent was observed.““@ Device efficiencies and parameters are not quantified in their report. Recently, Lee et al. reported an unusual changing of polarity of the transient photocurrent above a certain threshold of incident light pump intensity. (‘O’)The device, which consists of a sandwich structured microstripline configuration, shows a switch of the polarity of the photocurrent above Z, = 4.2 x lOI4photons/cm2. Furthermore, the sign of the photocurrent can be switched by irradiating another bias optical pump. This effect can potentially be useful for fast optical non-linear switching. Another interesting observation is the excitation profile of the steady state photocurrent which exhibit a change of the sign of the photocurrent at 2.3 eV by 0.3 V bias voltage. This optical photon energy where the internal field changes sign and thus is reflected by the phase of the observed photocurrent is a function of the applied bias voltage. The origin of this phenomenon remains unclear. Dember effect has been discussed in these devices, but the change in the sign of the photocurrent cannot be solely explained by this. 3.5. Fullerene thin jilm light emitting diodes: anomalous electroluminescence Since the HOMO-LUMO transition of Cm molecule is dipole forbidden, the solid state semiconductor band structure of C, resembles an indirect gap semiconductor. Thus, it is not expected that radiative recombination processes of charge carriers are favored. Surprisingly, there are reports of electroluminescence as well as photoluminescence which are highly non-linear and unusual. Yoshino and co-workers reported the observation of light emission from sandwich structures such as ITO/C,/Mg:In. (“‘) The intensity of the emitted light is proportional to the current at forward bias voltages and saturates at high current densities. Remarkable is the so called ‘white luminescence’, an unstructured, broad band emission ranging from 400 nm up to 1000 nm. A weak peak at around 530 nm was also observed.

Buckminsterfullerenes

155

in organic photoelectric devices

Another report on the broadband electroluminescent emission from fullerene films came from Werner et ~1.“~) Using gold and aluminium contacts as electrodes in a sandwich configuration, a broad band emission spectrum, extending from 400 to 1100 nm is observed. The spectrum has a maximum at about 920 nm. The spectral distribution of the electroluminescence is displayed in Fig. 21. Below a certain applied bias voltage (e.g. below a threshold current J,,) there is no detectable light emission. At higher currents, however, a non-linear emission sets in with a cubic increase in emission intensity upon increasing the current. The high current densities indicate a specific resistance of the C, film to be around 20 Rem. There is no resemblence between the two reported LED characteristics. The similarity of the electroluminescence described by Werner et al. and the photoluminescence spectrum of fullerenes in the highly excited state, however, clearly identifies the fullerene as the electroluminescent moeity. (“O) The observed cubic increase of the electroluminescence is similar to the non-linear behavior of the photoluminescence and the photoconductivity and therefore may be associated with the same physical phenomenon.“‘O~“” 4. CONCLUSION

AND OUTLOOK

The role of Buckminsterfullerene, C$,,, in conjugated, polymeric semiconductor devices is based on the strong electron accepting properties of C, upon photoexcitation. The above described photoinduced electron transfer phenomenon from conjugated polymers onto Cm 500

400

s 3

300

p’

2

2 ij

200

3 100

0

-

L

400

600

800 Wavelength

FIG. 21. The ‘white electroluminescence’ of Al/&/Au

1000

1200

(nm) Schottky devices (from Ref. 109)

156

N. S. Sariciftci

q

oouuo

CB

VB mmBw--

DADADA

DADADA

Photoinduced Electron Transfer

CB

VB

mmmm-m

DADADA

DADADA

FIG. 22. Band structural scheme of a heterostructure based on organic donor-acceptor successive layers. Photoinduced charge carriers can be stabilized at low temperatures rendering a photon flux dependent Fermi energy.

is ultrafast, yielding quantum efficiencies for the molecular charge transfer near unity! Furthermore, the charge separated state in these supramolecular composites is metastable, enabling the efficient collection of photoexcited charges in photovoltaic devices. Devices of quantum efficiencies around 20% have already been realized. Many other composite devices have been realized for photolithographic and xerographic purposes. This photophysical approach is promising for near future technological applications. Photoinduced electron transfer has also been observed in oligothiophene/C, composite films.(“2~“3)Since the vacuum sublimation method can be utilized for both oligomers of conjugated polymeric semiconductors”‘“‘20’ as well as Cm, it is conceivable that one can grow oligothiophene (donor) and C, (acceptor) multiple quantum well heterostructures by alternating the source in vacuum deposition process. These donor-acceptor (D-A-D-A-D-A-. . .) heterostructures are schematically displayed in Fig. 22. Such heterostructures should exhibit photoinduced excitations where the Fermi energy can be tuned by the intensity of the incident photon flux. Collective non-linear phenomena such as cooperative tunneling processes as a function of external bias field as well as incident photon flux can occur. Furthermore, the photoinduced charge generation and collection efficiency of such heterostructures is expected to scale with the number of heterojunction layers. This is potentially interesting for low noise, very high sensitivity photodiodes as a further improvement of the single layer heterojunction device discussed above. Recently, photoinduced polymerization of Cm thin films has been reported.02” Upon photoexposure the solubility of the C, thin film decreases considerably. This behavior is similar to negative photoresists in semiconductor manufacturing industry opening up the possibility of microstructures where Cso takes up the role of the active photoresist as well as the active semiconductor compound, simultaneously. This could open up an entire field of possibilities where microstructured semiconducting, metallic as well as superconducting arrays can be realized.

Buckminsterfullerenes

in organic photoelectric devices

157

With these prospects we will certainly be further excited by Buckminsterfullerenes in the future too; thus, Bucky rush continues. The prospects of organic n-electron semiconductors in general are unlimited due to the fact that flexibility of organic chemistry fused with interesting solid state physics aspects creates a vast space of parameters for design and modification of physical properties. Acknowledgements-The author gratefully acknowledges Alan J. Heeger and Fred Wudl for providing the possibilities, the support, as well ai their encouragementand feedback. This work is supported by the Department of Enernv No. DEFG0393ERl2138) ,. oroiect solar cells. We eratefullv acknowledee . _< (DOE \ ., for hieh efficiencv_ plastic our co-workers within the Institute for Polymers and Organic Solids at the University of Calrfornia, Santa Barbara, L. Smilowitz for photoinduced absorption studies, D. Braun and Zhang Chi for preparation of heterojunction devices, B. Kraabel for femtosecond photoinduced absorption studies, Changhee Lee for photoconductivity experiments, Kwanghee Lee for photoinduced infrared experiments, R. Janssen for photoinduced absorption experiments and C. Gettinger and R. Wu for preparation of P30T; from the Center for Quantized Electronic Structures at the University of California, Santa Barbara V. I. Srdanov for preparation of C, thin films, G. Wang, J. E. Bowers and G. D. Stucky as well as from UNIAX Corp. Santa Barbara, F. Klavetter for providing the MEH-PPV.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

20. 21. 22. 23.

24. 25. 26. 27. 28.

29. 30. 31. 32. 33. 34. 35. 36.

P. W. Stephens (ed.), Physics & Chemistry of Fullerenes, World Scienctific, Singapore (1993). J. Simon and J. J. Andre, Molecular Semiconductors, Springer, Berlin (1985). F. Gutmann and L. E. Lyons, Organic Semiconductors, Robert E. Krieger, Malabar (1981). J. E. Katon (ed.), Organic Semiconducting Polymers, Marcel Dekker, New York (1968). H. Meier, Organic Semiconductors, Verlag Chemie, Weinheim (1974). M. Pope and C. E. Swenberg, Electronic Processes in Organic Crystals, Clarendon Press, Oxford (1982). N. S. Sariciftci, L. Smilowitz, A. J. Heeger and F. Wudl, Science 258, 1474 (1992). N. S. Sariciftci, D. Braun, C. Zhang, V. Srdanov, A. J. Heeger, G. Stucky and F. Wudl, Appl. Phys. Left. 62, 585 (1993). N. S. Sariciftci, L. Smilowitz, A. J. Heeger and F. Wudl, Synth. Met. 59, 333 (1993). N. S. Sariciftci and A. J. Heeger, Inr. J. Mod. Phys. B8, 237 (1994). N. S. Sariciftci and A. J. Heeger, International Parent Publication no: WO 9405045, University of California, U.S.A. (1994). H. Shirakawa, E. J. Louis, A. G. MacDiarmid, C. K. Chiang and A. J. Heeger, Chem. Commun. 578 (1977). 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. Zert. 39, 1098 (1977). A. J. Heeger, S. Kivelson, J. R. Schrieffer and W.-P. Su, Rev. Mod. Phys. 60, 781 (1988). W. P. Su, J. R. Schrieffer and A. J. Heeger, Phys. Rev. Left. 42, 1698 (1979). W. P. Su, J. R. Schrieffer and A. J. Heeger, Phys. Rev. B22, 2099 (1980). W. P. Su, J. R. Schrieffer and A. J. Heeger, Phys. Reu. B28, 1138 (1983). H. Naarmann, Synth. Met. 17, 223 (1987). H. Naarmann and N. Theophilou, Synth. Met. 22, 1 (1987). T. Schimmel, W. Riess, J. Gmeiner, G. Denninger, M. Schwoerer, H. Naarmann and N. Theophilou, Solid State Commun. 65, 1311 (1988). T. Schimmel, G. Denninger, W. Riess, J. Voit, M. Schwoerer, W. Schoepe and H. Naarmann, Synth. Met. 28, Dll 91989). J. Chen, T. Chung, F. Moraes and A. J. Heeger, Solid State Commun. 53, 757 (1985). J. Chen and A. J. Heeger, Phys. Rev. B33, 1990 (1986). K. Kume, K. Mizuno, K. Mizoguchi, K. Nomura, J. Tanaka, M. Tanaka and H. Fuijimoto, Mol. Cryst. Liq. Cryst. 83, 49 (1982). Y. Cao and A. J. Heeger, Synth. Met. 52, 193 (1992). M. Reghu, Y. Cao, D. Moses and A. J. Heeger, Phys. Rev. B47, 1758 (1993). N. S. Sariciftci, A. J. Heeger and Y. Cao, Phys. Reu. LU9, 5988 (1994). K. H. Lee, A. J. Heeger and Y. Cao, Phys. Rev. B48, 14, 884 (1993). T. A. Skotheim (ed.), Handbook of Conducting Polymers, Marcel Dekker, New York (1986). Y. Lu, So/irons and Polarons in Conducting Polymers, World Scientific, Singapore (1988) J. M. Andre, J. Delhalle and J. L. Bredas, Quantum Chemistry Aided Design of Organic Polymers, World Scientific, New Jersey (1991). S. Kivelson and A. J. Heeger, Synth. Met. 22, 371 (1987). H. A. Pohl, In: Organic Semiconducting Polymers, p. 51, J. E. Katon (ed.), Marcel Dekker, New York (1968). C. 0. Yoon. M. Reghu, D. Moses and A. J. Heeger, Phys. Rev. B, submitted (1993). C. 0. Yoon. R. Menon, D. Moses and A. J. Heeger, Phys. REV. B49, 10851 (1994). R. D. McCullough, R. D. Lowe, M. Jayaraman and D. L. Anderson, J. Org. Chem. 58, 904 (1993).

158

N. S. Sariciftci

37. R. D. McCullough, S. Tristan-Nagle, S. P. Williams, R. D. Lowe and M. Jayaraman, J. Am. Chern. Sot. 115, 4910 (1993). 38. M. M. Bouman and E. W. Meijer, Polym. Prep. 35, 309 (1994). 39. Y. Cao, P. Smith and A. J. Heeger, Synrh. Met. 48, 91 (1992). 40. E. A. Silinsh, Organic Molecular Crystals, Springer, Berlin (1980). 41. L. Salem, Molecular Orbital Theory of Conjugated Systems, Benjamin, New York (1966). 42. M. A. Fox and M. Chanon (eds), Phoroinduced Electron Transfer, Elsevier, Amsterdam (1988). 43. A. H. Zewail (ed.), Photochemistry and Photobiology, Harwood Academic, Chur (1983). 44. F. L. Carter (ed.), Molecular Electronic Devices, Dekker, New York (1982). 45. A. Aviram (ed.), Molecular Electronics-Science and Technology, American Institute of Physics, New York (1992). 46. M. Mehring, In: Springer Series in Solid State Science, p. 242, H. Kuzmany, M. Mehring and S. Roth (eds), Springer, Berlin (1989). 47. M. R. Wasielewski, M. P. Niemcyk, W. A. Svec and E. B. Pewitt, J. Am. Chem. Sot. 106, 5043 (1984). 48. N. S. Sariciftci, A. Werner, A. Grupp, M. Mehring, G. Gotz, P. Bauerle and F. Effenberger, Mol. Phys. 75, 1259 (1992). 49. M. A. Fox, J. Wayne E. Jones and D. M. Watkins, Chemical & Engineering News March 15, (1993). 50. J. Guillet, Polymer Photophysics, Cambridge University Press, New York (1985). 51. S. E. Webber, Chem. Rev. 90, 1469 (1990). 52. Y. Wang, Nature 356, 585 (1992). 53. Y. Wang, R. West and C-H. Yuan, J. Am. Chem. Sot. 115, 3844 (1993). 54. R. J. Sension, A. Z. Szarka, G. R. Smith and R. M. Hochstrasser, Chem. Phys. L&r. 185, 179 (1991). 55. Y. Wang, J. Phys. Chem. %, 764 (1992). 56. Y. Wang and L.-T. Cheng, J. Phys. Chem 96, 1530 (1992). 57. R. G. Kepler and P. A. Cahill, Appl. Phys. Lett. 63, 1552 (1993). 58. P. V. Kamat, J. Am. Chem Sot. 113, 9705 (1991).

59. A. M. Rao, P. Zhou, K.-A. Wang, G. T. Hager, J. M. Holden, Y. Wang, W. T. Lee, A.-X. Bi, P. C. Eklund, D. S. Cornett, M. A. Duncan and I. J. Amster, Science 259, 955 (1993). 60. K. C. Hwang and D. Mauzerall, Nature 361, 138 (1993). 61. S. Morita, A. A. Zakhidov and K. Yoshino, Solid State Commun. 82, 249 (1992). 62. K. Yoshino, X. H. Yin, S. Morita, T. Kawai and A. A. Zakhidov, Solid State Commun. 85, 85 (1993). 63. K. Yoshino, X. H. Yin, K. Muro, S. Kiyomatsu, S. Morita, A. A. Zakhidov, T. Noguchi and T. Ohnishi, Jpn. J. Appl. Phys. 32, L357 (1993). 64. S. Morita, S. Kiyomatsu, M. Fukuda, A. A. Zakhidov, K. Yoshino, K. Kikuchi and Y. Achiba, Jpn. J. Appl. Phys. 32, Ll173 (1993). 65. S. Morita, S. Kiyomatsu, X. H. Yin, A. A. Zakhidov, T. Noguchi, T. Ohnishi and K. Yoshino, J. Appl. Phys. 74, 2860 (1993). 66. S. Morita, A. A. Zakhidov and K. Yoshino, Jpn. J. Appl. Phys. 32, L873 (1993).

67. K. Yoshino, T. Akashi, K. Yoshimoto, S. Morita, R. Sugimoto and A. A. Zakhidov, Solid State Commun. 90, 41 (1994). 68. L. Smilowitz, N. S. Sariciftci, R. Wu, C. Gettinger, A. J. Heeger and F. Wudl, Phys. Rev. B47, 13835 (1993). 69. P. M. Allemand, G. Srdanov, A. Koch, K. Khemani, F. Wudl, Y. Rubin, F. Diederich, M. M. Alvarez, S. J. Anz and R. L. Whetton, J. Am. Chem. Sot. 113, 2780 (1991). 70. B. Kraabel, C. H. Lee, D. McBranch, D. Moses, N. S. Sariciftci and A. J. Heeger, Chem. Phys. Leu. 213,389 (1993). 71. B. Kraabel, D. McBranch, N. S. Sariciftci, D Moses and A. J. Heeger, in preparation (1994). 72. S. M. Sze, Physics of Semiconductor Devices, John Wiley, New York (1981). 73. H. Yonehara and C. Pat, Appl. Phys. Left. 61, 575 (1992). 74. N. S. Sariciftci, L. Smilowitz, D. Braun, G. Srdanov, V. Srdanov, F. Wudl and A. J. Heeger, In: Proc. ICSM 92, Goteborg, Sweden, Synthetic Metals (1992). 75. Y. Yamashita, W. Takashima and K. Kaneto, Jpn. J. Appl. Phys. 32, L1017 (1993). 76. I. D. Parker, J. Appl. Phys. 75, 1656 (1994). 77. J. H. Burroughes, D. D. C. Bradley, A. R. Brown, R. N. Marks, K. Mackay, R. H. Friend, P. L. Burns and A. B. Holmes, Nature 347, 539 (1990). 78. D. Braun and A. J. Heeger, Appl. Phys. Left. 58, 1982 (1991). 79. D. Braun, A. J. Heeger and H. Kroemer, J. Electron. Mar. 20, 945 (1991). 80. D. Braun, G. Gustafsson, D. McBranch and A. J. Heeger, J. Appl. Phys. 72, 564 (1992). 81. Y. Ohmori, Y. Manda, H. Takahashi, T. Kawai and K. Yoshino, Jpn. J. Appl. Phys. 29, L837 (1990). 82. Y. Ohmori, M. Uchida, K. Muro and K. Yoshino, Solid State Commun. 80, 605 (1991). 83. Y. Ohmori, M. Uchida, K. Muro and K. Yoshino, Jpn. J. Appl. Phys. 30, Ll941 (1991). 84. H. Vestweber, A. Greiner, U. Lemmer, R. F. Mahrt, R. Richert, W. Heitz and H. Bassler, Adv. Mater. 4, 661 (1992). 85. G. Gustafsson, Y. Cao, G. M. Tracy, F. Klavetter, N. Colaneri and A. J. Heeger, Narure 357, 477 (1992). 86. G. Grem, G. Leditzky, B. Ullrich and G. Leising, Adv. Mater. 4, 36 (1992). 87. C. Zhang, S. Hager, K. Pakbaz, F. Wudl and A. J. Heeger, J. Electron. Mar. 22, 413 (1993). 88. N. C. Greenham, S. C. Moratti, D. D. C. Bradley, R. H. Friend and A. B. Holmes, Nature 366, 628 (1993). 89. G. Yu, C. Zhang and A. J. Heeger, Appl. Phys. Left. 64, 1540 (1994). 90. G. Yu, K. Pakbaz and A. J. Heeger, J. Electron. Mater. 23, 925 (1994).

Buckminsterfullerenes

in organic photoelectric devices

159

91. H. Antoniadis, B. R. Hsieh, M. A. Abkowitz, M. Stolka and S. A. Jehneke, Polymer Preprints 34,490 (1993). 92. G. Yu, K. Pakbaz and A. J. Heeger, Appl. Phys. Lerr. 64, 3422 (1994). 93. C. H. Lee, G. Yu, N. S. Sariciftci, D. Moses, K. Pakbaz, C Zhang, A. J. Heeger and F. Wudl, Phys. Rev. B48, 15425 (1993). 94. K. H. Lee, R. A. J. Janssen, N. S. Sariciftci and A. J. Heeger, Phys. Rev. B49, 5781 (1994). 95. C. S. Kuo, F. G. Wakim, S. K. Sengupta and S. K. Tripathy, Solid State Commun. 87, 115 91993). 96. R. F. Ziolo, US Parent No: US 52328i0, Xerox Corp., G.S.A. (1994). 97. J. Mort and M. A. Machonkin. U.S. Parenr No: U.S. 5178980. Xerox Corn., U.S.A. (19931. ~ ’ 98. J. Mort, M. Machonkin, I. Chin and R. Ziolo, Phil, Mug. Lek. 67, 77 (1693). 99. N. Aoki, German Parent No: DE 4321547, Nippon Oil Co., Japan (1994). 100. Y. Wang, International Parent Publicarion No: WO 93/02012, DuPont De Nemours and Company, U.S.A. (1993). 101. Y. Wang, International Parenr Publication No: WO 93108509, DuPont De Nemours and Company, U.S.A. (1993). 102. K. Shigematsu, Japanese Parenr No: 05088387, Idemitsu Kosan Co., Japan (1993). 103. A. F. Hebard, B. Miller, J. M. Rosamilia and W. L. Wilson, U.S. Parent No: 5171373, AT&T Bell Lab., U.S.A. (1992). 104. B. Miller, J. M. Rosamilia, G. Dabbagh, R. Tycko, R. C. Haddon, A. J. Muller, W. Wilson, D. W. Murphy and A. F. Hebard, J. Am. Chem. Sot. 113, 6291 (1991). 105. J. Mort, K. Okumura, M. Machonkin, R. Ziolo, D. R. Huffman and M. I. Ferguson, Chem. Phys. Lerr. 186, 281 (1991). 106. S. Curran, J. Callaghan, D. Weldon, E. Bourdin, K. Cazini, W. J. Blau, E. Waldron, D. McGoveran, M. Delamesiere, Y. Sarazin and C. Hogrel, In: Electronic Properties of Fullerenes, H. Kuzmany, J. Fink, M. Mehring, S. Roth (eds), Springer, Berlin (1993). 107. C. H. Lee, G. Yu, D. Moses, A. J. Heeger and V. I. Srdanov, Appl. Phys. Lerr. 65, 664 (1994). 108. M. Uchida, Y. Ohmori and K. Yoshino, Jpn. J. Appl. Phys. 30, L2104 (1991). 109. A. T. Werner, J. Anders, H. J. Byrne, W. K. Maser, M. Kaiser, A. Mittelbach and S. Roth, Appl. Phys. A57, 157 (1993).

110. H. J. Byrne, W. K. Maser, M. Kaiser, W. W. Riihle, A. Mittelbach and S. Roth, Chem. Phys. Lerr. 204, 461 (1993). 111. H. J. Byrne, W. K. Maser, M. Kaiser, L. Akselrod, J. Anders, W. W. Riihle, X. Q. Zhou, A. Mittelbach and S. Roth, Appl. Phys. A, in press. (1993). 112. R. A. J. Janssen, D. Moses and N. S. Sariciftci, J. Chem. Phys. in press (1994). 113. R. A. J. Janssen, D. Moses, N. S. Sariciftci and A. J. Heeger, In: Proc. Oprical Probes of Conjugated Polymers and Fullerenes, Salt Lake City, Utah, U.S.A., Z. V. Vardeny (eds) Mol. Cryst. Liq. Cryst., in press (1994). 114. G. Horowitz, D. Fichou, X. Peng, Z. Xu and F. Garnier, Solid Srare Comm. 72, 381 (1989). 115. D. Fichou, G. Horowitz, Y. Nishikitani and F. Gamier, Synrh. Met. 28, C723 (1989). 116. D. Fichou, G. Horowitz, Y. Nishikitani, J. Roncali and F. Gamer, Synrh. Met. 28, C729 (1989). 117. F. Garnier, G. Horowitz and D. Fichou, Synrh. Met. 28, C705 (1989). 118. F. Gamier, G. Horowitz, X. Peng and D. Fichou, Ado. Marer. 2, 592 (1990). 119. G. Horowitz, D. Fichou, X. Peng and F. Gamier, Synrh. Mer. 43, 1127 (1991). 120. D. Fichou, G. Horowitz and F. Gamier, In: Proc. Inrernarional Winrerschool on EIecrronic Properties of Polymers, Kirchberg, Austria, H. Kuzmany, M. Mehring, S. Roth, (eds) 107, Springer Series in Solid State Sci., 452 (1992). 121. A. M. Rao, P. Zhou, K. Wang, G. T. Hager, J. M. Holden, Y. Wang, W. Lee, X. X. Bi, P. C. Ecklund, D. S. Cornet, M. A. Duncan and I. J. Amster, Science 259, 955 (1993).