Charge separation type heterojunction structure and manufacturing method thereof Number:7,161,182 from the United States Patent and Trademark Office (PTO) owispatent A charge separation heterojunction structure which uses a fullerene polymer film as a part of its constituent materials and which may be used to produce a solar cellor a light emitting diode superior in durability, physical properties of electrons and economic merits. The heterojunction structure is such a structure in which an electron-donating electrically conductive high-polymer film and an electron-accepting fullerene polymer film are layered between a pair of electrodes at least one of which is light transmitting. In forming the layers, the fullerene polymer film is identified using in particular the Raman and Nexafs methods in combination so that upper layers are formed after identifying the polymer film.<H heterojunction, manufacturing, Charge, separati, 7161182.html, Patent, pto, 7161182

Title: Charge separation type heterojunction structure and manufacturing method thereof

Abstract: A charge separation heterojunction structure which uses a fullerene polymer film as a part of its constituent materials and which may be used to produce a solar cellor a light emitting diode superior in durability, physical properties of electrons and economic merits. The heterojunction structure is such a structure in which an electron-donating electrically conductive high-polymer film and an electron-accepting fullerene polymer film are layered between a pair of electrodes at least one of which is light transmitting. In forming the layers, the fullerene polymer film is identified using in particular the Raman and Nexafs methods in combination so that upper layers are formed after identifying the polymer film.

Patent Number: 7,161,182 Issued on 01/09/2007 to Ramm, et al.

Inventors: Ramm; Matthias (Kanagawa, JP), Ata; Masafumi (Kanagawa, JP)
Assignee: Sony Corporation (Tokyo, JP)
Research Institute of Innovative Technology for the Earth (Kyoto, JP)

Appl. No.: 10/314,858

Filed: December 9, 2002

Foreign Application Priority Data


Jun 25, 1999 [JP]			P11-179289
Jan 14, 2000 [JP]			P2000-005116

Current U.S. Class: 257/77 ; 257/12; 257/94

Current International Class: H01L 29/22 (20060101)

Field of Search: 257/77,12,22,94

References Cited [Referenced By]

U.S. Patent Documents


5331183	July 1994	Sariciftci et al.
5919429	July 1999	Tanaka et al.
5976477	November 1999	Isoda et al.
6017630	January 2000	Tanaka et al.
6113673	September 2000	Loutfy et al.
6117617	September 2000	Kanayama et al.
6455916	September 2002	Robinson

Foreign Patent Documents


829765	Sep., 1996	EP
05335614	Dec., 1993	JP
06029514	Feb., 1994	JP
06093258	Apr., 1994	JP
07147409	Jun., 1995	JP
8295505	Nov., 1996	JP
9309712	Dec., 1997	JP
11157819	Jun., 1999	JP

Other References

Rao A. M. et al., Photoinduced Polymerization of Solid C70 Films, Chemical Physics Letter, vol. 224, No. 1/2, pp. 106-112, (Jul. 8, 1994). cited by other .
Eklund, P.C. et al., Photochemical Transformation of C60 and C70 Films, Thin Solid Films, vol. 257, No. 2, pp. 185-203, (Mar. 1, 1995). cited by other .
Sariciftci N. S. et al., Semiconducting Polymer-Buckminsterfullerene Heterojunctions: Diodes, Photodiodes, and Photovoltaic Cells, Applied Physics Letters, American Institute of Physics, vol 62, No. 6, pp. 585-587, (Feb. 8, 1993). cited by other .
Sariciftci N. S. et al., Photovoltaic Cells Using Molecular Photoeffect at the Semiconducting Polymer/Buckminsterfullerene Heterojunctions, Extended Abstracts of the International Conference on Solid State Devices and Materials, pp. 781-783, (1993). cited by other.

Primary Examiner: Turner; Archene
Attorney, Agent or Firm: Sonnenschein, Nath & Rosenthal LLP

Parent Case Text

RELATED APPLICATION DATA

The present application claims priority to Japanese Application No. P11-179289, filed Jun. 25, 1999, and Japanese Application No. P2000-005116, filed Jan. 14, 2000, and is a divisional of U.S. application Ser. No. 09/597,247, filed Jun. 20, 2000 now abandoned, all of which are incorporated herein by reference to the extent permitted by law.

Claims

What is claimed is:

1. A charge separation heterojunction structure which comprises: an electrically conductive organic film and a fullerene polymer film layered between a light-transmitting electrode and a counter-electrode; said poly-fullerene film being a film formed by polymerization of fullerene molocules by one of polymerization on illumination of electron beam, plasma polymerization, micro-wave polymerization, electrolytic polymerization or a combination thereof.

2. The charge separation heterojunction structure according to claim 1 wherein said fullerene polymer film is contacted with said counter-electrode.

3. The charge separation heterojunction structure according to claim 1 wherein an active layer is interposed between said fullerene polymer film and said electrically conductive organic film.

4. The charge separation heterojunction structure according to claim 1 wherein said electrically conductive organic film has a covalent-electron system.

5. The charge separation heterojunction structure according to claim 1 wherein said light-transmitting electrode, electrically conductive organic film, fullerene polymer film and the counter-electrode are layered in this order on a substrate.

6. The charge separation heterojunction structure according to claim 1 wherein said light-transmitting electrode, fullerene polymer film, electrically conductive organic film and the counter-electrode are layered in this order on a substrate.

7. The charge separation heterojunction structure according to claim 4 wherein said electrically conductive organic film is at least one high molecular film of high molecular materials selected from the group consisting of polyvinyl carbazole, poly(p-phenylene) vinylene, polyaniline, polyethylene oxide, polyvinyl pyridine, polyvinyl alcohol, polythiophene, polyfluorene and polyparaphenylene, or a high molecular film obtained on polymerizing a derivative of at least one of monomer starting materials of the high molecular materials.

8. The charge separation heterojunction structure according to claim 1 wherein a dopant for controlling the electrically conductivity is added to said electrically conductive organic film.

9. The charge separation heterojunction structure according to claim 1 wherein said fullerene polymer film comprises a C60 polymer and/or a C70 polymer.

10. The charge separation heterojunction structure according to claim 1 wherein said poly-fullerene polymer film is formed by photopolymerization, polymerization on illumination of electron beam, plasma polymerization, micro-wave polymerization or electrolytic polymerization of fullerene molecules.

11. The charge separation heterojunction structure according to claim 1 comprising: a poly-fullerene polymer film obtained on polymerization of a vapor-deposited film of fullerene molecules on irradiation of electromagnetic waves.

12. The charge separation heterojunction structure according to claim 11 wherein said poly-fullerene polymer film is formed by polymerization of said vapor-deposited film vapor-deposited to a specified thickness.

13. The charge separation heterojunction structure according to claim 11 wherein an RF plasma, UV rays or an electron beam is illuminated as said electromagnetic waves.

14. The charge separation heterojunction structure according to claim 1 wherein said light-transmitting electrode and said counter-electrode are formed as thin films of a metal oxide or metal.

15. The charge separation heterojunction structure according to claim 14 wherein said light-transmitting electrode is a metal oxide obtained on doping an indium oxide with tin, or a thin film of gold, silver, platinum or nickel, and wherein said counter-electrode is a thin film of said metal oxide, aluminum, magnesium or indium.

Description

BACKGROUND OF THE INVENTION

1. Field of the Invention

This invention relates to a charge separation type heterojunction structure used in e.g., a solar cell or a light emitting diode and which has fullerene as a portion of a constituting material thereof. This invention also relates to a manufacturing method for the charge separation type heterojunction structure.

2. Description of Prior Art

Up to now, a silicon pn junction semiconductor etc has been extensively used in e.g., a solar cell or a light emitting diode. Of late, the energy conversion efficiency of the silicon pn junction semiconductor has been improved appreciably in comparison with that when the silicon pn junction semiconductor was initially devised.

Among the materials for the solar cell, there is e.g., titania in addition to silicon. Recently, however, fullerene, as a carbon compound, has attracted attention. The features of fullerene is hereinafter explained in connection with the discovery and the history of development thereof.

Fullerene is a series of carbon compounds composed only of carbon atoms, as is diamond or graphite. The existence of fullerene was confirmed in eighties. That is, it was found in 1985 in a mass analysis spectrum of a cluster beam by laser ablation of carbon. It was, however, five years later that the manufacturing method in reality was established. Specifically, a manufacturing method for fullerene (C.sub.60) by arc discharge of a carbon electrode was first found in 1990. Since then, fullerene is attracting notice as a carbonaceous semiconductor material (see Kratschmer, W., Fostiropoulos, K, Huffman D. R. Chem. Phys. Lett. 1990, 170, 167. Kratschmer, W. Lamb L. D., Fostiropoulod. K, Huffman, D. R. Nature 1990, 347, 354).

Fullerene is a spherical carbon C.sub.n (n=60, 70, 76, 78, 80, 82, 84, . . . ) which is a molecular aggregate resulting from spherical aggregation of an even number not less than 60 of carbon atoms. Representatives of the fullerenes are C.sub.60 with 60 carbon atoms and C.sub.70 with 70 carbon atoms. Of these, the C.sub.60 fullerene is of a polyhedral structure termed truncated-icosahedron obtained from an icosahedron by truncating each of the twelve vertices. Hence, each vertex is replaced by a pentagon. Thus, the C.sub.60 fullerene has a molecular structure of what may be termed a soccer ball type in which its 60 apices are all occupied by carbon atoms. On the other hand, C.sub.70 has what may be termed a rugby ball type molecular structure.

In a C.sub.60 crystal, C.sub.60 molecules are arranged in a face-centered cubic structure. It has a band gap of approximately 1.6 eV and may be deemed as a semiconductor. In an intrinsic state, it has an electrical resistivity of approximately 10.sup.7 .OMEGA.cm. It has a vapor pressure of approximately 1 m Torr at 500.degree. C. and, on sublimation, is capable of vapor depositing a thin film. Not only C.sub.60 but other forms of the fullerene are readily vaporized in vacuum or under reduced pressure and hence are able to yield an evaporated film easily.

However, the molecules of fullerene forms, such as C.sub.60 or C.sub.70, the most mass-producible, are of zero dipole moment, such that evaporated films produced therefrom are fragile in strength, because only the van der Waal's force acts between its molecules. Thus, if the evaporated film is exposed to air, molecules of oxygen or water tend to be diffused and intruded into the gap between the fullerene molecules (FIG. 2), as a result of which the evaporated film is not only deteriorated in structure but adverse effects may be occasionally produced in its electronic properties. This fragility of the fullerene poses a problem in reference to device stability when applying the fullerene to fabrication of a thin-film electronic device.

For overcoming the weak points the fullerene polymer film, described above, the method of producing a so-called fullerene polymer consisting in polymerizing fullerene molecules has been proposed. Typical of these methods is a method of forming a fullerene polymer film by light excitation [see (a) Rao, A. M., Zhou, P, Wang., K. A, Hager., G. T., Holden, J. M., Wang, Y., Lee, W. T., Bi, X, X., Eklund, P. C., Cornet, D. S., Duncan, M. A., Amster, J. J. Science 1993, 256995, (b) Cornet, D. C., Amster I. J., Duncan, M. A., Rao A. M., Eklund P. C., J. Phys. Chem. 1993, 97,5036, (c) Li. J., Ozawa, M., Kino, N, Yoshizawa, T., Mitsuki, T., Horiuchi, H., Tachikawa, O; Kishio, K., Kitazawa, K., Chem. Phys. Lett. 1994, 227, 572].

In these methods, in which light is illuminated on a previously formed evaporated fullerene film, numerous cracks tend to be formed in the film surface due to volumetric contraction produced on polymerization, so that produced films are problematic in strength. Moreover, it is extremely difficult to form a uniform thin film of a large surface area.

It has also been known to apply pressure or heat to fullerene molecules or to cause collision of fullerene molecules against one another. It is however difficult to produce a thin film, even though it is possible to form a film (see, for a molecule collision method, (a) Yeretzian, C., Hansen, K., Diedrich, F., Whetten, R. L., Nature 1992, 359,44, (b) Wheten, R. L., Yeretzian, C., Int. J. Multi-layered optical disc. Phys. 1992, B6,3801, (c) Hansen, K., Yeretzian, C., Whetten, R. L., Chem. Phys. Lett. 1994, 218,462, and (d) Seifert, G., Schmidt, R., Int. J. Multi-layered optical disc. Phys. 1992, B6,3845; for an ion beam method, (a) Seraphin, S., Zhou, D., Jiao, J. J. Master. Res. 1993, 8,1995, (b) Gaber, H., Busmann, H. G., Hiss, R., Hertel, I. V., Romberg, H., Fink, J., Bruder, F., Brenn, R. J. Phys. Chem., 1993, 97,8244; for a pressure method, (a) Duclos, S. J., Brister, K., Haddon, R. C., Kortan, A. R., Thiel, F. A. Nature 1991, 351,380, (b) Snoke, D. W., Raptis, Y. S., Syassen, K. 1 Phys. Rev. 1992, B45, 14419, (c) Yamazaki, H., Yoshida, M., Kakudate, Y., Usuda, S., Yokoi, H., Fujiwara, S., Aoki, K., Ruoff, R., Malhotra, R., Lorents, D. J., Phys. Chem. 1993, 97, 11161, and (d) Rao, C. N. R., Govindaraj, A., Aiyer, H. N., Seshadri, R. J. Phys. Chem. 1995, 99,16814).

Noteworthy as a fullerene polymerization method or film-forming method, which should take the place of the above-enumerated fullerene polymerization methods, is the plasma polymerization method or the micro-wave (plasma) polymerization method, previously proposed by the present inventors in e.g., Takahashi, N., Dock, H. or in Matsuzawa, N., Ata M. J., Appl. Phys. 1993, 74,5790. The fullerene polymer film, obtained by these methods (see FIGS. 3 and 4), are thin films produced by polymerization of the fullerene molecules through an electronic excited state. It is appreciably increased in strength in comparison with the evaporated thin fullerene film, dense and high in pliability. Since the fullerene polymer film is scarcely changed in its electronic properties in vacuum or in air, it may be premeditated that its dense thin-film properties effectively suppress diffusion or intrusion of oxygen molecules into the inside of the film. In reality, generation of fullerene polymer consisting the thin film by these methods may be demonstrated by the time-of-flight mass spectrometry.

Irrespective of the type of the plasma method, electron properties of a fullerene polymer film possibly depend appreciably on its polymerization configuration. In reality, the results of mass spectrometry of the C.sub.60 polymer film, obtained by the micro-wave plasma method, bear strong resemblance to those of the C.sub.60 argon plasma polymer thin film, previously reported by the present inventors [see Ata, M., Takahashi, N., Nojima, K., J. Phys. Chem. 1994, 98, 9960, Ata, M., Kurihara, K., Takahashi, J. Phys., Chem., B., 1996, 101, 5].

The structure of the fullerene polymer may be estimated by the pulse laser excited time-of-flight mass spectrometry (TOF-MS). In general, there is known a matrix assist method as a method for non-destructive measurement the high molecular polymer. However, lacking the solvent capable of dissolving the fullerene polymer, it is difficult to directly evaluate the actual molecular weight distribution of the polymer. Even with the mass evaluation by Laser Desorption Ionization Time-of-Fight Mass Spectroscopy (LDITOF-MS), it is difficult to make correct evaluation of the mass distribution of an actual fullerene polymer due to the absence of suitable solvents or to the reaction taking place between C.sub.60 and the matrix molecule.

The structure of the C.sub.60 polymer can be inferred from the profile of a dimer or the peak of the polymer of LDITOF-MS, as observed in the ablation of such a laser power as not to cause polymerization of C.sub.60. For example, LDITOF-MS of a C.sub.60 polymer film, obtained with a plasma power of e.g., 50 W, indicates that the polymerization of C.sub.60 molecules is most likely to take place through a process accompanied by loss of four carbon atoms. That is, in the mass range of a dimer, C.sub.120 is a minor product, whilst C.sub.116 is produced with the highest probability.

According to semi-empiric C.sub.60 dimer calculations, this C.sub.116 may be presumed to be D2h symmetrical C.sub.116 shown in FIG. 10. This may be obtained by C.sub.58 recombination. It is reported that this C.sub.58 is yielded on desorption of C.sub.2 from the high electronic excited state including the ionized state of C.sub.60 [(a) Fieber-Erdmann, M., et al., Phys. D. 1993, 26,308 (b) Petrie, S. et al., Nature 1993, 356,426 and (c) Eckhoff, W. C., Scuseria, G. E., Chem. Phys. Lett. 1993, 216,399].

If, before transition to a structure comprised of two neighboring five-membered rings, this open-shell C.sub.58 molecules are combined with two molecules, C.sub.116 shown in FIG. 10 is produced. However, according to the notion of the present inventors, it is after all the [2+2] cycloaddition reaction by the excitation triplex mechanism in the initial process of the C.sub.60 plasma polymerization. The reaction product is shown in FIG. 9. On the other hand, the yielding of C.sub.116 with the highest probability as mentioned above is possibly ascribable to desorption of four sp.sup.3 carbons constituting a cyclobutane of (C.sub.60).sup.2 yielded by the [2+2] cycloaddition from the excited triplet electronic state of C.sub.60 and to recombination of two C.sub.58 open-shell molecules, as shown in FIG. 6.

If a powerful pulsed laser light beam is illuminated on a C.sub.60 fine crystal on an ionization target of TOF-Ms, as an example, polymerization of fullerene molecules occurs through the excited electronic state, as in the case of the micro-wave plasma polymerization method. At this time, ions of C.sub.58, C.sub.56 etc are also observed along with peaks of the C.sub.60 photopolymer.

However, since no fragment ions, such as C.sub.58.sup.2+ or C.sup.2+ are observed, direct fragmentation from C.sub.60.sup.3+ to C.sub.58.sup.2+ and to C.sup.2+, such as is discussed in the literature of Fieber-Erdmann, cannot be thought to occur in this case. Also, if C.sub.60 is vaporized in a C.sub.2 F.sub.4 gas plasma to form a film, only addition products of fragment ions of F or C.sub.2F.sub.4 of C.sub.60 are observed in the LDITOF-MS, while no C.sub.60 polymer is observed. Thus, the LDITOF-MS, for which no C.sub.60 polymer is observed, has a feature that no C.sub.58 nor C.sub.56 ions are observed. These results of observation support the fact that C.sub.2 loss occurs through a C.sub.60 polymer.

The next problem posed is whether or not the C.sub.2 loss is directly caused from 1, 2-(C.sub.60).sub.2 produced by the [2+2] cycloaddition reaction shown in FIG. 6. Murry and Osawa et al proposed and explained the process of structure relaxation of 1,2-(C.sub.60) 2 as follows [(a) Murry, R. L. et al, Nature 1993, 366,665, (b) Strout, D. L. et al, Chem. Phys. Lett. 1993, 214,576, Osawa, E, private letter].

Both Murry and Osawa state that, in the initial process of structure relaxation of 1,2-(C.sub.60).sub.2, C.sub.120 (d) of FIG. 13 is produced through C.sub.120 (b) of FIG. 11, resulting from cleavage of the 1,2-C bond, having the maximum pinch of the cross-linked site, from C.sub.120 (c) of FIG. 12 having the ladder-like cross-linking by Stone-Wales transition (Stone, A. J., Wales, D. J., Chem. Phys. Lett. 1986, 128,501, (b) Saito, R. Chem. Phys. Lett. 1992, 195,537). On transition from C.sub.120 (c) of FIG. 12 to C.sub.120 (b) of FIG. 11, energy instability occurs. However, on further transition from C.sub.120 (c) of FIG. 12 to C.sub.120 (d) of FIG. 13, the stabilized state is restored.

Although it is not clear whether the nC.sub.2 loss observed in the polymerization of C60 by plasma excitation directly occurs from 1,2-(C.sub.60) thought to be its initial process or after certain structure relaxation thereof, it may be premeditated that the observed C.sub.118 assumes the structure shown in FIG. 14 by desorption of C.sub.2 from C.sub.120 (d) of FIG. 13 and recombination of dangling bonds. Also, C.sub.116 shown in FIG. 15 is obtained by desorption of two carbon atoms of the ladder-like cross-linking of C.sub.118 of FIG. 14 and recombination of bonds. Judging from the fact that there are scarcely observed odd-numbered clusters in the dimeric TOF-MS, and from the structural stability, it may be presumed that the loss in C.sub.2 is not produced directly from 1,2-(C.sub.60) 2, but rather that it is produced through C.sub.120 (d) of FIG. 13.

Also, Osawa et al states in the above-mentioned literature that D5d symmetrical C.sub.120 structure is obtained from C.sub.120 (a) through structure relaxation by multi-stage Stone-Wales transition. However, insofar as the TOF-MS of the C.sub.60 polymer is concerned, it is not the structure relaxation by the multi-stage transition reaction but rather the process of structure relaxation accompanied by C.sub.2 loss that governs the formation of the polymer by plasma irradiation.

In a planar covalent compound in general, in which a .pi.-orbital crosses the .sigma.-orbital, spin transition between 1(.pi.-.pi.*)-3(.pi.-.pi.*) is a taboo, while it is allowed if, by vibration-electric interaction, there is mixed the .sigma.-orbital. In the case of C.sub.60, since the .pi.-orbital is mixed with the .sigma.-orbital due to non-planarity of the .pi. covalent system, inter-state crossing by spin-orbital interaction between 1(.pi.-.pi.*) and 3(.pi.-.pi.*) becomes possible, thus producing the high photochemical reactivity of C.sub.60.

The plasma polymerization method is applicable to polymerization of C.sub.70 molecules. However, the polymerization between C.sub.70 molecules is more difficult to understand than in that between C.sub.60. Thus, the polymerization is hereinafter explained in as plain terms as possible with the aid of numbering of carbon atoms making up C.sub.70.

The 105 C--C bonds of C.sub.70 are classified into eight sorts of bonds represented by C(1)-C(2), C(2)-C(4), C(4)-C(5), C(5)-C(6), C(5)-C(10), C(9)-C(10), C(10)-C(11) and C(11)-C(12). Of these, C(2)-C(4) and C(5)-C(6) are of the same order of double bond performance as the C.dbd.C in C.sub.60. The .pi.-electrons of the six members of this molecule including C(9), C(10), C(11), C(14) and C(15) are non-localized such that the C(9)-C(10) of the five-membered ring exhibit the performance of the double bond, while the C(11)-C(12) bond exhibits single bond performance. The polymerization of C.sub.70 is scrutinized as to C(2)-C(4), C(5)-C(6), C(9)-C(10) and C(10)-C(11) exhibiting the double-bond performance. Meanwhile, although the C(11)-C(12) is substantially a single bond, it is a bond across two six-membered rings (6,6-ring fusion). Therefore, the addition reaction performance of this bond is also scrutinized.

First, the [2+2] cycloaddition reaction of C.sub.70 is scrutinized. From the [2+2] cycloaddition reaction of these five sorts of the C--C bonds, 25 sorts of dimers of C.sub.70 are produced. For convenience of calculations, only nine sorts of the addition reactions between the same C--C bonds are scrutinized. Table 1 shows heat of the reaction (.DELTA.Hf0(r)) in the course of the process of yielding C.sub.140 from C.sub.70 of two molecules of the MNDO/AN-1 and PM-3 levels.

In the table, .DELTA.Hf0(r)AM-1 and .DELTA.Hf.degree. (r)PM-3 means calculated values of the heat of reaction in case of using parameterization of the MNDO method which is a semi-empirical molecular starting method by J. J. Stewart.

TABLE-US-00001 TABLE 1 cluster .DELTA.Hf0(r) .DELTA.Hf0(r) bond (reference (kcal/mol) (kcal/mol) length drawing) AM-1 PM-3 cross-linking (.ANG.) C140(a) -34.63 -38.01 C(2)-C(2'), C(4)-C(4') 1544 (FIG. 15) C(2)-C(4), C(2)-C(4') 1607 C140(b) -34.33 -38.00 C(2)-C(4'), C(4)-C(2') 1544 (FIG. 16) C(2)-C(4), C(2')-C(4') 1607 C140(c) -33.94 -38.12 C(5)-C(5'), C(6)-C(6') 1550 (FIG. 17) C(5)-C(6), C(5')-C(6') 1613 C140(d) -33.92 -38.08 C(5)-C(6'), C(6)-C(5') 1551 (FIG. 18) C(5)-C(6), C(5')-C(6') 1624 C140(e) -19.05 -20.28 C(9)-C(9'), C(10)-C(10') 1553 (FIG. 19) C(9)-C(10), C(9')-C(10') 1655 C140(f) -18.54 -19.72 C(9)-C(10'), C(10)-C(9') 1555 (FIG. 20) C(9)-C(10), C(9')-C(10') 1655 C140(g) +3.19 -3.72 C(10)-C(10'), C(11)-C(11') 1559 (FIG. 21) C(10)-C(11), C(10')-C(11') 1613 C140(h) +3.27 -3.23 C(10)-C(11'), C(11)-C(10') 1560 (FIG. 22) C(10)-C(11), C(10')-C(11') 1613 C140(i) +64.30 +56.38 C(11)-C(11'), C(12)-C(12') 1560 (FIG. 23) C(11)-C(12), C(11)-C(12') 1683

In the above Table, C.sub.140 (a) and (b), C.sub.140 (c) and (d), C.sub.140 (e) and (f) and C.sub.140 (g) and (h) are anti-syn isomer pairs of the C(2)-C(4), C(5)-C(6), C(9)-C(10) and C(10)-C(11) bonds, respectively. In the addition reaction between C(11) and C(12), only D2h symmetrical C.sub.140 (i) is obtained. These structures are shown in FIGS. 15 to 23. Meanwhile, an initial structure of a C.sub.70 polymer by the most stable [2+2] cycloaddition is shown in FIG. 14.

From this Table 1, no energy difference is seen to exist between the anti-syn isomers. The addition reaction between the C(2)-C(4) and C(5)-C(6) bonds is as exothermic as the addition reaction of C60, whereas that between the C(11)-C(12) is appreciably endothermic. Meanwhile, the C(1)-C(2) bond is evidently a single bond. The heat of reaction of the cycloaddition reaction in this bond is +0.19 and -1.88 kcal/mol at the AM-1 and PM-3 level, respectively, which are approximately equal to the heat of reaction in C.sub.140 (g) and (h). This suggests that the cycloaddition reaction across the C(10) and C(11) cannot occur thermodynamically. Therefore, the addition polymerization reaction across the C.sub.70 molecules occurs predominantly across the C(2)-C(4) and C(5)-C(6), whereas the polymerization across the C(9)-C(10) bonds is only of low probability, if such polymerization takes place. It may be premeditated that the heat of reaction across the C(11)-C(12), exhibiting single-bond performance, becomes larger than that across the bond C(1)-C(2) due to the appreciably large pinch of the cyclobutane structure of C.sub.140 (i), in particular the C(11)-C(12) bond. For evaluating the effect of superposition of the 2p2 lobe of sp2 carbon neighboring to the cross-linking bondage at the time of [2+2] cycloaddition, the values of heat generated in the C.sub.70 dimer, C.sub.70-C.sub.60 polymer and C.sub.70H.sub.2 were compared. Although detailed numerical data are not shown, it may be premeditated that the effect of superposition can be safely disregarded across C.sub.140 (a) to (h), insofar as calculations of the MNDO approximate level are concerned.

The mass distribution in the vicinity of the dimer by the LDITOF-MS of the C.sub.70 polymer film indicates that dimers of C.sub.116, C.sub.118 etc are main products. Then, scrutiny is made into the structure of C.sub.136 produced on desorbing four carbon atoms making up cyclobutane of a dimer (C.sub.70) 2, as in the process of obtaining D2h-symmetrical C.sub.116 from C.sub.60 and recombining remaining C.sub.68. These structures are shown in FIGS. 28 to 36. Table 2 shows comparative values of the generated heat (.DELTA.Hf0) of C.sub.136.

In Table 2, H.DELTA.f0 AM-1, .DELTA.Hf0 PM-3, cross-linking and the binding length are the same as those of Table 1.

TABLE-US-00002 TABLE 2 cluster .DELTA.Hf0(r) .DELTA.Hf0(r) bond (reference (kcal/mol) (kcal/mol) length drawing) AN-1 PM-3 cross-linking (.ANG.) C136(a) -65.50 -61.60 C(1)-C(8'), C(3)-C(5') 1.351 (FIG. 24) C(5)-C(3'), C(8)-C(1') 1.351 C136(b) -64.44 -61.54 C(1)-C(3'), C(3)-C(1') 1.351 (FIG. 25) C(5)-C(8'), C(8)-C(5') 1.351 C136(c) 0 0 C(4)-C(13'), C(7)-C(10') 1.352 (FIG. 26) C(10)-C(7'), C(13)-C(4') 1.352 C136(d) +0.09 +0.11 C(4)-C(7'), C(7)-C(4') 1.351 (FIG. 27) C(10)-C(13'), C(13)-C(10') 1.354 C136(e) +112.98 +102.89 C(5)-C(8'), C(8)-C(5') 1.353 (FIG. 28) C(11)-C(14'), C(14)-C(11') 1.372 C136(f) +69.47 +59.44 C(5)-C(14'), C(14)-C(5') 1.358 (FIG. 29) C(11)-C(8'), C(8)-C(11') 1.352 C136(g) -3.74 -9.20 C(5)-C(15'), C(15)-C(5') 1.344 (FIG. 30) C(12)-C(9'), C(9)-C(12') 1.352 C136(h) +2.82 -5.30 C(5)-C(9'), C(9)-C(5') 1.372 (FIG. 31) C(12)-C(15'), C(15)-C(12') 1.334 C136(i) +98.50 +84.36 C(13)-C(10'), C(15)-C(16') 1.376 (FIG. 32) C(10)-C(13'), C(16)-C(15') 1.376

It is noted that C.sub.136 (a) to (i) are associated with C.sub.140 (a) to (i), such that C(2) and C(4), which formed a cross-link at C.sub.140(a), have been desorbed at C.sub.136(a). It is noted that carbon atoms taking part in the four cross-links of C.sub.136(a) are C(1), C(3), C(5) and C(8), these being SP2 carbon atoms. Among the dimers shown in Table 1, that estimated to be of the most stable structure at the PM-3 level is C.sub.140(c). Therefore, in Table 2, .DELTA.Hf0 of C.sub.136(c), obtained from C.sub.140(c), is set as the reference for comparison. It may be seen from Table 2 that the structures of C.sub.136(a) and C.sub.136(b) are appreciably stabilized and that C.sub.136(e), C.sub.136(f) and C.sub.136(i) are unstable. If the calculated values of .DELTA.Hf0 of per a unit carbon atom of the totality of C.sub.140 and C.sub.136 structures are evaluated, structure relaxation in the process from C.sub.140 to C.sub.136 only take place in the process from C.sub.140(a) and (b) to C.sub.136(a) and (b). Thus, the calculations of the MNDO approximation level suggest that, in the C.sub.70 cross-link, not only are the sites of the [2+2] cycloaddition of the initial process limited to the vicinity of both end five-membered rings traversed by the main molecular axis, but also is the cross-link structure of the .pi.-covalent system, such as C.sub.136, limited to C.sub.136 obtained from the dimer of C.sub.70 by the cycloaddition reaction across C(2)-C(4) bond. The molecular structure of more stable C136, yielded in the process of relaxation of the structure shown in FIG. 13, is shown in FIG. 14.

The polymer film of C.sub.60 shows semiconducting properties with band gap evaluated from temperature dependency of the dark current being of the order of 1.5 to 2 eV. The dark conductivity of the C.sub.60 polymer film obtained with the micro-wave power of 200 W is on the order of 10.sup.-7 to 10.sup.-8 S/cm, whereas that of the C.sub.70 polymer film obtained for the same micro-wave power is not higher than 10.sup.-13 S/cm, which is approximate to a value of an insulator. This difference in the electrical conductivity of the polymer films is possibly attributable to the structures of the polymer films. Similarly to the sole cross-link bond in which two-molecular C.sub.60 is in the state of open-shell radical state, the cross-link of a dimer of 1,2-C(60) due to [2+2] cycloaddition reaction of FIG. 1 is thought not to contribute to improved electricaly conductivity. Conversely, the inter-molecular cross-link, such as C.sub.116, forms the .pi.-covalent system, and hence is felt to contribute to improved electrically conductivity. The cross-link structures of C.sub.118, C.sub.114 and C.sub.112, now under investigations, are thought to be a .pi.-covalent cross-link contributing to electrically conductivity.

It may be contemplated that the electrical conductivity usually is not increased linearly relative to the number of electrically conductive cross-links between fullerene molecules, but is changed significantly beyond the permeation limit at a certain fixed number. In the case of C.sub.70, the probability of the [2+2] cycloaddition reaction is presumably lower than that in the case of C.sub.60, while the structure relaxation to the electrically conductive cross-linked structure such as that from C.sub.140 to C.sub.136 can occur only on specified sites. In light of the above, the significant difference in electrically conductivity between the two may possibly be attributable to the fact that, in the C.sub.60 polymer film, the number of cross-links contributing to electrically conductivity is large and exceeds the permeation limit, whereas, in the case of C.sub.70, the permeation limit is not exceeded because of the low probability of polymerization and limitation of formation of electrically conductive cross-links.

Taking into account the discovery of the fullerene molecule, its evaporated film and fullerene polymer film and the mechanism of polymerization thereof, discussed in the foregoing, we return to the discussion of the solar cell referred to in the beginning part of the present specification.

The material fullerene has latent possibility of yielding a solar cell improved both economically and as to physical characteristics. As a matter of fact, several solar cells having fullerene as its constituent material have so far been proposed (see JP Patents Nos.9656473, 95230248 and 99325116, U.S. Pat. No. 5,171,373 and WO 9405045).

However, the solar cells, hitherto proposed, are common in exploiting the fullerene evaporated film, so that the above-mentioned problem attributable to the fragility of the evaporated film, in particular the durability or physical properties of electrons, as yet remains unsolved.

Meanwhile, the fullerene polymer film, belonging to the fullerene system as does the above-mentioned evaporated film, exhibits sufficient durability due to its superior physical properties such as freeness from oxygen diffusion into the polymer bulk material. However, it has scarcely been attempted up to now to use the material as a constituent material for fabrication of the solar cells.

This may possibly be attributable to the circumstances that the industrial fullerene polymerization technique has been developed only recently. In addition, the fact that the method of definitely identifying the fullerene polymer film by a non-destructive technique has not been established possibly needs to be taken into consideration.

As means for clarifying the interconnection of carbon skeletons in a carbonaceous compound, there is also known a method such as a nuclear magnetic resonance method. However, insofar as carbonaceous thin film, such as a fullerene polymer film, is concerned, difficulties are encountered in measurement due to failure in definite observation of the pattern of free induction attenuation depending on electrical conductivity and to transverse relaxation to nuclear spin by dangling unpaired spin.

Moreover, the nuclear magnetic resonance method is not suited as means for monitoring structural changes in the carbonaceous thin film material due to difficulties encountered in magic angle spin of an individual sample.

OBJECT AND SUMMARY OF THE INVENTION

It is therefore an object of the present invention to provide a charge separation type heterojunction structure in which, by establishing a method for non-destructively identifying a fullerene polymer film and by applying the polymer film to a layered structure, it is possible to produce a solar cell etc which is improved not only in durability and electronic physical properties but also in various physical properties.

In one aspect, the present invention provides a fullerene polymer obtained on polymerizing a vapor-deposited film of fullerene molecules by irradiation of electromagnetic waves.

In another aspect, the present invention provides a method for manufacturing a charge separation heterojunction structure including the steps of forming a light-transmitting electrode, forming an electrically conductive organic film, forming a fullerene polymer film and forming a counter-electrode, wherein the steps of forming constituent layers other than the fullerene polymer film is carried out after first identifying the fullerene polymer film.

In the heterojunction structure of the present invention, in which an electrically conductive organic film as an electron donor, and a fullerene polymer film, as an electron acceptor, are layered between a pair of electrodes, at least one of which is light transmitting. Therefore, the heterojunction structure of the present invention finds application in solar cells or light emitting diodes. Since the fullerene polymer film is used as a part of the constituent material, the heterojunction structure of the present invention is particularly superior in durability and electronic physical properties as compared to the case of employing a fullerene deposited film. A vapor-deposited film tends to lose its desirable characteristics in about one day in the evaluation in atmosphere. Conversely, a polymerized vapor-deposited film is scarcely changed in characteristics even after one month.

Moreover, if the heterojunction structure is applied to a solar cell, it is possible to produce a thin film which is lower in cost, more light weight and flexible than a film used in a conventional silicon pn junction solar cell.

In the manufacturing method of the present invention, in which the respective constituent layers of the charge separation heterojunction structure can be formed without difficulties, and the fullerene polymer film can be identified by using the Raman and Nexafs methods in combination, the variable evaluation including that on the structure and the polymerization degree of the fullerene polymer film, amorphization, oxidation and dielectric breakdown by application of a high voltage, can be realized precisely non-destructively. The fullerene polymer film can be accurately identified based on the results of evaluation on the fullerene polymer film, so that the targeted heterojunction structure can be fabricated reliably. Moreover, the results of the evaluation can be used for controlling its physical properties.

BRIEF DESCRIPTION OF THE DRAWINGS

FIGS. 1A and 1B schematic views showing the molecular structure of fullerene, where FIG. 1A shows the molecular structure of C.sub.60 and FIG. 1B shows that of C.sub.70.

FIG. 2 shows the structure of a C.sub.60 evaporated film.

FIG. 3 shows an illustrative structure of a C.sub.60 polymer.

FIG. 4 shows an illustrative structure of a C.sub.60 polymer.

FIG. 5 shows a dimeric structure of a C60 molecule.

FIG. 6 shows a dimeric structure of another C.sub.60 molecule.

FIG. 7 shows another dimeric structure of another C.sub.60 molecule [C.sub.120(b)].

FIG. 8 shows still another dimeric structure of another C.sub.60 molecule [C.sub.120(c)].

FIG. 9 shows yet another dimeric structure of another C.sub.60 molecule [C.sub.120(d)].

FIG. 10 shows a structure of a C.sub.118 molecule.

FIG. 11 shows a structure of a C.sub.116 molecule.

FIG. 12 shows a numbering system of a C.sub.70 molecule.

FIG. 13 shows a dimeric structure of a C.sub.70 molecule.

FIG. 14 shows another dimeric structure of a C.sub.70 molecule.

FIG. 15 shows still another dimeric structure of a C.sub.70 molecule.

FIG. 16 shows still another dimeric structure of a C.sub.70 molecule.

FIG. 17 shows still another dimeric structure of a C.sub.70 molecule.

FIG. 18 shows still another dimeric structure of a C.sub.70 molecule.

FIG. 19 shows still another dimeric structure of a C.sub.70 molecule.

FIG. 20 shows still another dimeric structure of a C.sub.70 molecule.

FIG. 21 shows still another dimeric structure of a C.sub.70 molecule.

FIG. 22 shows still another dimeric structure of a C.sub.70 molecule.

FIG. 23 shows still another dimeric structure of a C.sub.70 molecule [C.sub.140(i): D2h symmetrical].

FIG. 24 shows still another dimeric structure of a C.sub.70 molecule [C.sub.136(a)].

FIG. 25 shows still another dimeric structure of a C.sub.70 molecule [C.sub.136(b)].

FIG. 26 shows still another dimeric structure of a C.sub.70 molecule [C.sub.136(c)].

FIG. 27 shows still another dimeric structure of a C.sub.70 molecule [C.sub.136(d)].

FIG. 28 shows still another dimeric structure of a C.sub.70 molecule [C.sub.136(e)].

FIG. 29 shows still another dimeric structure of a C.sub.70 molecule [C.sub.136(f)].

FIG. 30 shows still another dimeric structure of a C.sub.70 molecule [C.sub.136(g)].

FIG. 31 shows still another dimeric structure of a C.sub.70 molecule [C.sub.136(h)].

FIG. 32 shows still another dimeric structure of a C.sub.70 molecule [C.sub.136(i)].

FIGS. 33A and 33B show an illustrative heterojunction structure of the present invention, where FIG. 33A is a schematic cross-sectional view showing a simple hetero structure and FIG. 33B is a schematic cross-sectional view showing a simple hetero structure.

FIGS. 34C and 34D show schematic cross-sectional views of another simple heterojunction structure, where FIG. 34C shows a simple heterojunction structure and FIG. 34D shows a double hetero structure.

FIG. 35 shows an apparatus for yielding fullerene molecules by arc discharge.

FIG. 36 shows an apparatus for producing a fullerene polymer film by the plasma polymerization method and by the vapor deposition film electro-magnetic wave illumination method.

FIG. 37 shows an apparatus for producing a fullerene polymer film by the micro-wave polymerization method.

FIG. 38 is a schematic cross-sectional view showing a film-forming process of a fullerene polymer film by the vapor deposition film electro-magnetic wave illumination method.

FIG. 39 is a schematic cross-sectional view of the fullerene vapor deposition device.

FIG. 40 is a schematic cross-sectional view showing a plasma polymerization device for polymerization of the fullerene polymer film.

FIG. 41 shows an apparatus for producing a fullerene polymer film by the electrolytic polymerization method.

FIGS. 42A and 42B show the state of separation of electrons and holes in a heterojunction of the heterojunction structure according to the present invention, where FIG. 42A shows the state in the absence of a step and FIG. 42B shows the state in the presence of a step.

FIG. 43 shows the Raman spectrum of amorphous carbon.

FIG. 44 shows Fermi level of an ITO thin film.

FIG. 45 shows the results of PES measurement of a polythiophene thin film.

FIG. 46 shows diode characteristics of the polythiophene thin film.

FIG. 47 shows the results of Raman spectroscopic measurement of a C60 polymer film in comparison with those of the graphitic carbon.

FIG. 48 shows the measured results of photoelectron emission of a fullerene thin film.

FIG. 49 is a C 1s spectral diagram of XPS of the fullerene polymer film according to the present invention.

FIG. 50 is a C 1s peak distribution diagram of the fullerene polymer film.

FIG. 51 is a spectrum diagram showing a shake-up satellite area of the fullerene polymer film.

FIG. 52 shows the XPS valence band of the fullerene polymer film according to the present invention.

FIG. 53 is a TOF-MS spectral diagram of the fullerene polymer film obtained by the plasma processing.

FIG. 54 is a TOF-MS spectrum diagram of the fullerene polymer film obtained by the plasma processing.

FIG. 55 shows a band structure of an example of a heterojunction structure according to the present invention.

FIG. 56 shows V-I characteristics of the heterojunction structure shown in FIG. 55.

FIG. 57 shows V-I characteristics of another his according to the present invention.

FIG. 58 shows V-I characteristics of a layered structure corresponding to the heterojunction structure.

FIG. 59 shows V-I characteristics of the layered structure in which the material of the counter-electrode of the layered structure is changed.

FIG. 60 shows the Raman spectrum of a C.sub.60 polymer film obtained by the plasma polymerization method.

FIG. 61 shows the Raman spectrum of a C.sub.60 polymer film obtained by changing the conditions of the plasma polymerization method.

FIG. 62 shows a Raman spectrum of a fullerene polymer film according to the present invention.

FIG. 63 shows a Raman spectrum of a C.sub.60 vapor-deposited film.

FIG. 64 shows a Raman spectrum of graphit.

FIG. 65 shows a Raman spectrum of a C.sub.60 polymer film, obtained by an argon plasma polymerization method, and a C.sub.60 polymer film, obtained by an electrolytic polymerization method.

FIG. 66 shows Raman spectra of C.sub.60 polymer film obtained by the plasma polymerization method.

FIG. 67 shows Raman spectra of C.sub.60 polymer film obtained by the plasma method in case the plasma power is kept constant and the pressure is changed.

FIG. 68 shows a Raman spectrum of a C.sub.60 vapor-deposited film.

FIG. 69 shows Raman spectrum of a C.sub.60 polymer film when the pressure is changed.

FIG. 70 shows Raman spectrum of a C.sub.60 polymer film when the pressure is changed further.

FIG. 71 shows Raman spectrum of a C.sub.60 polymer film obtained by the plasma polymerization method under a constant power in case the pressure is changed.

FIG. 72 shows the Raman spectrum of the C.sub.60 polymer film in case the plasma power is changed.

FIG. 73 shows the Nexafs spectrum of the C.sub.60 vapor-deposited film.

FIG. 74 shows the Nexafs spectrum of the C.sub.60 plasma polymer film.

FIG. 75 shows the expanded portion .pi.-antibonding orbital region for each sample of FIGS. 73 and 74.

FIG. 76 shows the Nexafs spectrum of a C60 polymer film of a sample of FIG. 61.

FIG. 77 shows the Nexafs spectrum of oxygen K edge of a thin carbon film of the sample of FIG. 76.

FIG. 78 shows the relationship between the spectrum of Nexafs method and electronic transitions.

FIG. 79 shows V-I characteristics of a heterojunction structure having an active layer interposed between the electrically conductive high polymer film and a fullerene polymer film, according to the present invention.

FIG. 80 shows photoelectron emitting spectra of phthalocyanine film.

FIG. 81 shows the relationship between the photoelectron energy and absorption coefficients of the phthalocyanine film.

FIG. 82 shows V-I characteristics when the counter-electrode in the heterojunction structure of FIG. 79 has been changed.

FIG. 83 shows absorption spectrum of ITO-tetrathiafulvalene-C60 polymer structure.

FIG. 84 shows V-I characteristics of another heterojunction structure according to the present invention.

FIG. 85 shows an illustrative cross-sectional structure of a heterojunction structure according to the present invention and particularly shows a five-layer structure.

FIG. 86 is a schematic view showing the structure of a film-forming apparatus for a heterojunction structure.

FIG. 87 shows V-I characteristics on light illumination of the heterojunction structure.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

A charge separation heterojunction structure of the present invention is preferably of the above-described layered structure in which the fullerene polymer film contacts with the counter-electrode.

Also, inclusive of this case, an active layer is preferably interposed as a carrier generating layer between the fullerene polymer film and the electrically conductive organic film.

According to the present invention, a substrate may preferably be provided on an outer surface side, that is a surface exposed to atmosphere, of each electrode.

As a typical heterojunction structure of the present invention, it is preferably of such a structure in which a light-transmitting electrode 2, such as ITO (indium tin oxide) comprised of a transparent substrate 1 of silicon or glass, an electrically conductive high polymer film 3 of e.g., polythiophene, a fullerene polymer film 4 forming a heterojunction with the electrically conductive high polymer film, and a counter-electrode 5 of, for example, aluminum, are layered in this order on the transparent substrate 1, as shown in FIG. 33A. An active layer 6 of, for example, carbon nanotube or phthalocyanine, is preferably interposed as a carrier generating layer between the electrically conductive high polymer film 3 and the fullerene polymer film 4, as shown in FIG. 33B. The film thicknesses of the electrically conductive high polymer film 3, active layer 6 and the fullerene polymer film 4 are preferably 0.1 to 50 nm and more preferably 5 to 20 nm, hereinafter the same.

It is noted that charge separation is also possible in a heterojunction structure in which the electrically conductive high polymer film 3 and the fullerene polymer film 4 are interchanged with each other, as shown in FIGS. 34C and 34D, such that this structure is also comprised in the scope of the present invention.

The electrically conductive organic film is electron-donating and is preferably comprised of p-type electrically conductive high molecular material containing a covalent .pi.-electron system. Preferred examples of the polymers include those of polyvinyl carbazole, poly (p-phenylene)-vinylene, polyaniline, polyethylene oxide, polyvinylpyridine, polyvinyl alcohol, polythiophene, polyfluorene, polyparaphenylene and derivatives of these constituent monomers.

Meanwhile, these electrically conductive organic films may be admixed with known dopants, such as sulfuric acid radicals, for controlling their electrically conductivity.

The fullerene polymer film operates as an electron accepting thin film and preferably composed of a C.sub.60 polymer and/or C.sub.70 polymer, such as those shown for example in FIGS. 5 to 11, in FIGS. 13 to 32 and in FIGS. 3 and 4. However, the polymers are not limited to these examples. In comparison with the vapor-deposited fullerene film, shown in FIG. 2, this fullerene polymer film features tight bonding among fullerene molecules by covalent bonds.

The active layer, as an optionally provided layer, is a carrier generating layer, and is formed of dyes having a .pi.-electron system, metal complexes, electrically conductive high polymer materials, fullerene molecules, chemically modified derivatives thereof, single- or multi-layered carbon nanotubes, either alone or in combination. The dyes may be enumerated by cyanine dyes, phthalocyanine, metal complexes thereof, porphyrin and metal complexes thereof. As the material of the light-transmitting electrode, the ITO (indium oxides doped with tin) is generally preferred. However, thin films of gold, silver, platinum or nickel may also be used.

The materials of the counter-electrode may be enumerated by one or more of metals, such as aluminum, magnesium, indium, alloys thereof, or ITO.

The manufacturing method of the present invention for producing the charge separation heterojunction structure basically includes a step of forming a light-transmitting electrode, a step of forming an electrically conductive organic film, a step of forming a fullerene polymer film and a step of forming a counter-electrode. The sequence of these steps is, however, arbitrary, on the condition that the fullerene polymer film is to be identified first before proceeding to the formation of the remaining constituent layers. The step of mounting a substrate on the light-transmitting electrode or the counter-electrode or the step of interposing an active layer may be added as appropriate.

For this identification, the Raman method and the Nexafs method, both being non-destructive spectroscopic method, are preferably used in combination. If one of these methods is omitted, the identification cannot be executed satisfactorily.

Among items of evaluation of the specified identification operations, there are a fullerene polymer structure, polymerization degree, amorphization, oxidation and insulation destruction by impression of a high voltage. The results of the evaluation may be used not only for identification of the fullerene polymer film but also for controlling physical properties, such as control of polymerization conditions.

In the step of forming the light-transmitting electrode, the routine practice is forming the electrode on a substrate, rather than forming it alone. The electrode material, such as ITO, is formed on the substrate by techniques such as vapor deposition or sputtering.

If thin film of a stable metal, such as gold, is used in place of ITO film, it is crucial to form the thin film to a thin thickness on the substrate to provide light transmittance. The shape or the pattern of the light-transmitting electrode may be freely selected by known means, such as mask.

The electrically conductive organic film or the fullerene polymer film is formed on the light transmitting electrode. Meanwhile, in case the fullerene polymer film is formed on the light-transmitting electrode, the procedure may be simply reversed from that employing the electrically conductive organic film, the corresponding explanation is omitted for simplicity.

During this forming step, a vapor-deposited film or a plasma polymerization film of an organic low molecular compound, exhibiting electron donating properties, is formed on the light-transmitting electrode.

That is, if a monomer of the high molecular material or an organic low molecular compound containing .pi.-electrons is vaporized and the gas thus yielded is irradiated with a high frequency plasma of a lower energy, UV rays or an electron beam, an electrically conductive organic film can be produced on the light-transmitting electrode.

The vapor-deposited film or the plasma polymerization film of the .pi.-covalent organic low molecular material has electrically conductivity at least of the order of 10.sup.-9 S/cm. Since the fullerene polymer film, as later explained, operates as an electron accepting thin film, the low molecular organic vapor-deposited film or the plasma polymerization film needs to operate as an electron donating thin film.

The low molecular organic compounds may be enumerated by a .pi. covalent low molecular material, such as ethylene or acetylene, cata-condensation organic compounds, such as benzene, naphthalene or anthracene, peri-condensation aromatic compounds, such as perillene or coronene, and derivatives of these compounds as to hetero atoms, such as nitrogen, oxygen or sulfur. It is noted that oxygen, sulfur, selenium or tellurium can be built as hetero atom into an organic skeleton, however, since these atoms normally furnish two electrons to the .pi.-electron system, there are furane or thiophene as a hetero cyclic compound of oxygen or sulfur which proves a .pi.-electron system with e.g., benzene. If one of these elements is built into a six-membered ring or two of the elements are built into a five-membered ring, the .pi.-electron system is present in excess amount in view of the 4n+2 rule so that the resulting compound is strongly electron-donating. Typical of the strong electron-donating compounds is tetrathiafullvalene. The vapor-deposited film or the plasma polymerization film of this strongly electro-donating organic compound forms a heterojunction with an electron-accepting fullerene polymer film as later explained to induce charge separation by light induction more effectively.

The materials for forming the electrically conductive organic film, as polymers, may be enumerated by high molecular materials or derivatives thereof, in addition to the above-mentioned polyvinyl carbazole and polythiophene, these may being used alone or in combination.

These materials are poly (3-alkylthiophene), poly[2-methoxy-5-(2'-ethylhexoxi)-p-phenylene]-vinylene, poly [2-methoxy-5-(2'-ethylhexoxi)-1,4-paraphenylene vinylene, poly (3-alkylthiophene), poly (9,9-dialkylfluorene), polyparaphenylene, poly (2,5-diheptiloxy-1,4-phenylene), polyphenylene, polyaniline, poly (p-phenylene), polyethylene oxide, poly (2-vinyl pyridine and poly(vinyl alcohol). It is also possible to execute polymerization by illumination of a high frequency plasma of a lower energy, or UV-rays, X-rays or electron rays, in a gaseous atmosphere of these high molecular compounds or an organic compound containing the .pi.-electron system, to produce a highly electrically conductive organic thin film.

On the electrically conductive organic film, produced as described above, a fullerene polymer film is formed by the following procedure.

First, the fullerene molecules, as a starting material, those of C.sub.60, C.sub.70 and higher-order fullerene may be used, either singly or in combination. Most preferred are the C.sub.60 fullerene, the C.sub.70 fullerene or mixtures thereof. In addition, the fullerene of higher orders, such as C.sub.70, C.sub.78, C.sub.80, C.sub.82, C.sub.84 and so forth may be contained therein.

These fullerene molecules may be manufactured by an arc discharge method of a carbon electrode, using an apparatus shown for example in FIG. 35.

In a reaction vessel 8 of the present apparatus, there are mounted a pair of carbon electrodes, connected to an AC or DC source 9, such as counter-electrodes 10a, 10b formed of graphite. After evacuating the reaction vessel 8 by a vacuum pump through an outlet 11b by an exhaust pump, low-pressure inert gas, such as helium or argon, is introduced via an inlet 11a so as to be charged into the reaction vessel 8.

The ends of the counter-electrodes 10a, 10b are arranged facing each other with a small gap in-between, and a predetermined current and voltage are applied from the DC source 9 to maintain the state of arc discharge across the ends of the counter-electrodes 10a, 10b for a predetermined time.

By this arc discharge, the counter-electrodes 10a, 10b are vaporized so that soot is gradually deposited on a substrate 12 mounted on the inner wall surface of the reaction vessel 8. If this amount of soot deposited is increased, the reaction vessel 8 is cooled and the substrate 12 is taken out, or the soot is recovered using a sweeper.

From this soot, the fullerene such as C.sub.60 or C.sub.70 may be extracted using a .pi.-electron based solvent, such as toluene, benzene or carbon disulfide. The yielded fullerene, obtained in this stage, is termed crude fullerene, which may be applied to column chromatography to separate C.sub.60 and C.sub.70 as purified separate products.

The resulting fullerene molecules are used as a starting material in the film-forming process of the fullerene polymer. Among the polymerization or film-forming methods, there are, for example, a photopolymerization method, an electron beam illumination method, a plasma polymerization method, a micro-wave polymerization method) and an electrolytic polymerization method.

Photo Polymerization Method

In this polymerization method, an apparatus including a reaction chamber capable of being maintained at a reduced pressure or in vacuum, heating means, such as resistance heating means, for vaporizing the fullerene molecules, and illumination means for illuminating the light, such as ultraviolet beam, through the window of the reaction chamber, is used. A fullerene polymer film is formed on the substrate as fullerene is evaporated and illumination of ultraviolet light is continued for a predetermined time. At this time, the fullerene molecules are excited by light and polymerized through the excited state.

It is