Effects of annealing on the electrical conductivity of C60 films

Thin

Solid Films 261 (1995) 266-270

Effects of annealing on the electrical conductivity Jinlong

Gong,

Fangqing

Zhang,

Yahong

Li, Guobin

of C,, films

Ma, Guanghua

Chen

Department of Physics, Lanzhou University, Lanzhou 730000, People’s Republic of China

Received 10 August

1994; accepted

31

January 1995

Abstract We report the annealing effects on the electrical conductivity of C,,, films. Films sublimed using C,, powder from the same batch were used. The studies show that the C,, films have clear semiconducting behavior in the temperature range extending from room temperature to 473 K. The room-temperature conductivity is in the range of lo-’ to 10-7(Rcm)P’. From the measurements of conductivity versus time when the film is maintained at 473 K, we explained that the increase of conductivity before 2.5 h of annealing is the result of a decrease of the unstable hcp phase and an increase of the number of connected crystallites, the decrease of conductivity after 2.5 h of annealing is due to the decrease of the number of connected crystallites. The decrease in the number of connected crystallites makes the intercrystalline potential barrier higher, and thus decreases the conductivity. The reason for the increase of the activation energy is the decrease of defects between connected crystallites. These defects may introduce states in the energy gap of CeOfilms. The deviation of the conductivity from straight lines in the high-temperature region in the a versus l/r plots are due to unstable structural phases which transform toward a stable fee phase in the annealing process.

Keywords:

Annealing;

Carbon; Electrical properties and measurements

1. Introduction Since the initial discovery of buckminsterfullerene(C,,) in 1985 [l] and the subsequent development of the carbon arc method for producing macropcopic quantities of this material [2], a large amount of effort has been attracted to the understanding and application of this new allotrope of carbon [3]. The discovery of superconducting properties of alkali-metal-doped C,, fullerides has directed the researchers toward the detailed characterization of their electrical properties [47]. Comparatively much less information is available on the electrical properties of undoped fullerenes [8-171. Crystalline C,, was reported to be a molecular semiconductor [6, 18, 191 in which C,, molecules sit on the sites of a fee lattice with the lattice parameter a = 1.417 nm [20]. Because of the special structure and chemical properties of C,, molecules, it is possible to introduce atoms of clusters inside the C,, cage, to add atoms or clusters outside the cage, and to substitute a few carbon atoms on the spherical shell by different species. Various new compounds can be synthesized by 0040-6090/95/$9.50 0 1995 SSDI 0040-6090(95)06542-3

Elsevier Science S.A. All rights

reserved

chemical modification for applications in different areas. However, a neccessary step is the detailed investigation of the intrinsic properties of pure C,, fullerene, and the study of how they are affected when the material is exposed to various experimental factors. Except for the electrical properties, many investigations have been made on pure C,, fullerene. Different results have been obtained from different studies owing to the experimental details. A series of different techniques including theoretical predictions indicated a range of 1.5-2.3 eV for the fundamental band gap of C,, films [l 1, 121. The earliest study on the electrical properties reported an upper limit of 10-5(Qcm)-’ for the roomtemperature conductivity [ 111. The temperature-dependent conductivity in lightly doped alkali-metal C,,(K., C,,,) films has also been observed [21]. Mort et al. reported 10-‘4(Rcm)’ for the room-temperature conductivity for the C&,/C,, films [lo]. Electrical properties of single-crystal C,, have been measured by different researchers [14- 161, and a room-temperature electrical conductivity of 10 ~ ‘@cm) ~ ’ was obtained [14]. Hamed et al. [12] reported that the activation

J. Gong et ul.

: Thin Solid Films 261 (I 995) .Wp-771)

energies and room-temperature dark conductivities were in the range 0.54-0.58 eV and 1Om-h to lo- ‘(Rem) ’ for the oxygen-free C,, polycrystalline films. They also found a high sensitivity of the conductivity on oxygen contamination and on illuminaLion. This indicates that the oxygen quickly permeates the whole depth of the C,, films and produces a disorder potential that leads to the localization of the electronic states at the edges of highest-occupied-molecular-orbital- and lowest-unoccupied-molecular-orbital-derived bands. In addition, oxygen may act as an efficient trap for electrons in the conduction band of C,, . The electrical properties are seriously affected by oxygen, and illumination speeds up the reaction of oxygen with C,,, which results in the increase of defect states in the films. So excluding oxygen from the films seems very important in the measurements of electrical properties of C,,, films. Here we report our measurements on the electrical dark conductivity of C,,, films. We study the annealing effects on the electrical transport properties. The annealing temperature was higher than 443 K, so the oxygen effects can be neglected [12].

2. Experimental Carbon soot containing C,, and other forms of carbon was produced by arc discharge between two highpurity graphite rods in a 13.3 kPa He atmosphere. &,, C,,, and a small amount of higher fullerenes were extracted from the soot by toluene, and then separated by chromatography. Then the obtained C,, powder was washed by ethyl ether and recrystallized in toluene to obtain a purity higher than 99.5%. The C,, powder was dryed at 393 K in a dynamic vacuum of lo-’ Pa for 10 h, followed by another 10 h at 573 K. The C,, powder was then stored protected from light. For the preparation of the C,, films, a small amount of C,,, powder was placed in a molybdenum boat and kept at 5733623 K in a vacuum of 7 x lop4 Pa for several hours, and then the boat was heated to 823-873 K to evaporate the C,,, onto an unheated glass substrate 8 cm apart from the source. All the films used for the measurements were prepared with the same batch of C,, powder. Before the measurements, silver electrodes having a length of 0.5 cm and a gap of 0.3 cm were evaporated onto the C,, films, and Cu wires were attached to one edge of the electrodes with silver paste. We measured that silver electrodes exhibited an ohmic contact up to a 100 V applied field. The conductivity was measured in the temperature range 298-473 K using heating and cooling rates of 2 K min’. The film thickness was 460 nm, as determined by surface profilometry. X-ray diffraction (XRD) indicated

261

that the film was polycrystalline with a very complex structure, including unstable phases such as an hcp phase, monoclinic phases and an amorphous phase. But the major phase was fee.

3. Results and discussion On the first heating, the conductivity varies in a wide range as shown in Ref. [12], and the conductivity is of the order of 10 ‘3(Qcm)- ’ at 298 K. A curve in place of a straight line for the logarithms of conductivity versus inverse temperature was obtained. This is because of the existence of oxygen in the film. So the curve is not characteristic of pure C,,,. Oxygen changes the electronic states of the C,,, band. Mort et al. [lo] varied the temperature from 330 K to 230 K. In this temperature range, the oxygen adsorbed in C,,, films cannot be desorbed or expelled out. They obtained a conductivity of 10p’4(Rcm) ~’ at room temperature. Such a low conductivity could be related to oxygen adsorbed in the films. After one hour of annealing at 473 K, we measured the temperature-dependent conductivity on cooling (Fig. I, curve a); then measured the temperature-dependent conductivity on heating again. The two curves are identical for temperatures lower than 400 K. This indicates that oxygen was expelled out of the film and the curves should be the characteristic of pure C,,, if we

-‘I---

-2 -

-3 -

t

_4-

$ “0 5 -5

-6

6

Fig. I. The temperature dependence of the conductivity (r for C,,, film after: curve a, I h (0); curve b, 2 h (” ); curve c, 3.5 h (KY);curve d. 5 h ( n ); curve e, 8 h (A); curve f. IO h (A I. of holding at 473 K measured on cooling.

J. Gong et al. / Thin Solid Films 261 (1995) 266-270

268

Table 1 CJ~), 298 K conductivity annealing. Annealine

I 2 3.5 5 8 10

time (h)

a(298 K) and

0{,(km) 3.72 1.11 1.50 3.80 1.10 8.30

x x x x x x

- ’ 10’ 104 103 IO’ 104 lo7

activation

energy

E,

(r(298 K)(Rcm)-’

E, (ev)

2.86 6.90 8.30 5.60 9.50 9.50

0.48(2) 0.48(4) 0.48(8) 0.52(2) 0.59(5) 0.61(2)

x x x x x x

IO-’ IO-’ l0-h 1O-6 lo-7 IO-’

after

neglect the effects of residual oxygen. The straight line in the log cr versus l/T plot indicates that the conductivity obeys the law: 0 = rrOexp( - E, /kT) at temperatures below 400 K and the material shows clear semiconducting behavior, where rrOis the conductivity prefactor, k the Boltzmann constant, T the absolute temperature, and E, the activation energy. We can obtain the conductivity at 298 K, a(298 K) = 2.86 x 10-S(Rcm)-‘. This value is close to that of Ref. [I 11, but higher than that of Ref. [12] by two orders of magnitude. At temperatures higher than 400 K, the conductivity shows a downward curvature deviating from the straight line. We will provide an explanation of this phenomenon later. After annealing in vacuum at 473 K for 2 h, 3.5h, 5 h, 8 h and 10 h respectively, we measured the temperature-dependent conductivities, with the results shown in Fig. 1, curves b, c, d, e, and f respectively. Besides Fig. 1, curve a, Fig. 1, curve b also shows downward curvature at higher temperatures. After 3.5 h of annealing (Fig. 1, curve c) this downward curvature disappeared. The plot of logarithm of conductivity versus inverse temperature becomes a straight line in the whole temperature range after 5 h or more of annealing. We believe that the reason for the high-temperature curvature is the unstable structure or defects. The structural instability makes the conductivity vary at high temperature, and then the dependence of conductivity on temperature shows downward character at high temperatures. From the straight lines in Fig. 1, we can obtain activation energies (E,), co values and conductivities at 298 K after different annealing times as shown in Table 1. Fig. 2 shows the conductivity at 298 K (a(298 K)) as a function of annealing time. The conductivity at 298 K increases with annealing time before 2.5 h, and then decreases. The conductivity after 10 h of annealing is two orders of magnitude lower than that after 2 h. The conductivity at 298 K after 10 h of annealing is close to that mentioned in Ref. [12], which is the result measured for fee polycrystalline C,, films. So we think that our films may experience a transition from unstable phases to the stable fee polycrystalline phase. Because

-4 c

0

1

2

3

4

5

6

7

8

9

10

11

time (hour) Fig. 2. The conductivity at 298 K for the C,,, film (a(298 K)) as a function of annealing time.

of the weak van der Waals force between C,, molecules, the structure of the C,, film can be easily relaxed on annealing with the fee polycrystalline phase increasing and unstable phases decreasing, XRD results supported this assumption. The details concerning the XRD results will be published elsewhere. In the XRD spectrum of our C,, film before annealing, we found that besides the fee phase, there also exists an hcp phase as well as a monoclinic phase, and the fee (111) and (333) diffraction lines have extraordinarily strong intensities in comparison with the (220), (311) difraction lines. After annealing at 473 K for 10 h, the (11 l), (220) (311) diffraction lines have comparable intensities and the (333) line becomes very weak. These results also show that the hcp and monoclinic phases disappeared. This indicated that the fee phase is a more stable phase than the hcp and monoclinic phases at high temperature. The XRD results of the C,, film after annealing at 473 K for 2 h show that only the unstable hcp phase disappeared. Comparatively very little change of the intensity of the fee and monoclinic diffraction lines were observed. The XRD results after annealing at 453 K and 463 K for 10 h are similar to that at 473 K for 2 h. According to the above structural information, we see clearly that there are more crystallites exposing the (111) face in the film before annealing than after, these crystallites may be connected together by the monoclinic phase and play an important role in the conductivity of C,, films. The truth of the monoclinic phase will be studied by scanning tunneling microscopy or atomic force microscopy. We think that there is an intercrystalline phase between the fee crystallites. The connected crystallites have a lower intercrystalline potential barrier. Thus the film can have a higher conductivity. After annealing for 10 h, the number of crystallites exposing different faces are comparable. The

269

J. Gong et al. 1 Thin Solid Films 261 (1995) 266-270

0.60

_

.

3

10-j I

I

I

012

I

3

I

I

45 time

Fig, 3. The activation

energy

I

6

I

78

1

1

9

0

I

10

11

(how)

E, as a function

of annealing

time.

crystallites may not be connected together owing to the disappearance of the monoclinic phase and give a higher intercrystalline potential barrier and lower conductivity of the film. In Fig. 3, activation energy E, is plotted as a function of annealing time. The activation energy varies slowly before 2.5 h and after 8 h. Between 2.5 h and 8 h, the activation energy increases steeply. The three regions can be understood as follows. When annealing time is less than 2.5 h, the film has a mixed phase structure, and the unstable hcp phase decreases with increasing annealing time. When annealing time is longer than 8 h, the film is in the fee phase and has almost the same number of crystallites exposing the (ill), (311) (220) faces, respectively. Between the two regions, the film is in the relaxation process with crystallites exposing the (111) face and the monoclinic phase decreasing. The intercrystalline connection may introduce some type of defect (such as stacking fault). These defects introduce defect gap states which result in the smaller activation energy E,. Annealing makes the number of connected crystallites decrease and thus decreases the gap states, so the activation energy increases. In order to understand the phenomena above, Fig. 4 illustrates the conductivity as a function of time when the film is maintained at 473 K. Before 2.5 h, the conductivity increases. After 2.5 h, the conductivity decreases and tends toward a constant about 2 x lOme’(Rem) ~’ for 20 h annealing. We have measured the coductivity as a function of time at 453 K and 463 K. In 10 h of observation, the conductivity increases to a constant, we did not observe any decrease in conductivity. Only when the temperature is above 473 K does the conductivity decrease with time. Based on the experimental results, including the XRD results: firstly, we can see that before 2.5 h the unstable hcp phase decreases with time, and the number of fee

2

4

6

8

10

12

14

16

18

20

time (hour) Fig. 4. Time dependence of the conductivity temperature was maintained at 473 K.

of the C,,, film when the

crystallites connected together increases. This makes the conductivity increase. When the annealing time is longer than 2.5 h, the unstable monoclinic phase decreases and the number of connected fee crystallites decreases, the intercrystalline potential barrier becomes higher, thus the conductivity decreases; secondly, as XRD results of films annealing at 453 K and 463 K are similar to that at 473 K for 2 h, we may suggest that annealing at 453 K and 463 K only decreases the unstable hcp phase. Only when the annealing temperature is higher than 473 K can the change of the fee and monoclinic structure take place in a manner in which fee crystallites exposing the (111) face and monoclinic phase decrease. From Fig. 4, we can obtain an explanation for the deviation of the conductivity from the straight lines in Fig. l(curves a and b). Before 2.5 h of annealing, the conductivity increases; cooling makes the conductivity decrease. The two tendencies slow the rate of decrease of conductivity when the temperature cools down, and in the high-temperature region, conductivity shows a downward curvature. After 2.5 h of annealing the conductivity decreases; cooling also makes conductivity decrease. The total effect should speed up the rate of decrease of the conductivity, and in the high-temperature region, conductivity should show upward curvature. But within our measurement error range the conductivity is a straight line. This is because when the temperature cools down below 473 K, the fee and monoclinic structure cannot change any further. When the annealing time is less than 2.5 h, there exists an unstable hcp phase in the film, which may transform toward a stable fee phase when the temperature is higher than about 400 K. From the discussion above, we can see clearly that the deviation of the conductivity from the straight line is the result of unstable structure in the films.

270

J. Gong rt al. / Thin Solid Films 261 (1995) 266-270

Hamed et al. [12] suggested an explanation for the downward curvature on the g versus I/r curve. They thought of that as the result of a large temperaturedependent volume expansion coefficient for this Van der Waals solid, for which the primary effect of a lattice expansion on the electronic structure is the narrowing of the electronic bands and the widening of the energy gaps. But after 20 h of annealing, their plot of 0 versus l/T was a straight line, so it seems unreasonable to explain the deviations of conductivity from straight lines in the high-temperature region using lattice expansion. We can suggest an explanation for this phenomenon as follows, when the thickness of films is greater than 2000 &I, the C,, molecules continuously deposited interact with the substrate weakly, and the motion cannot be restricted by the substrate. Unstable structures can easily take place. These unstable structures transform toward a stable structure under the conditions of 180 “C deposition. The conductivity will increase and the plots of 0 versus l/T show a downward curavature.

4. Conclusions Studies of the annealing effects on the electrical conductivity and the temperature-dependent conductivity of C,, films show that the films have clear semi-conducting behavior in the temperature range extending from room temperature to 473 K, and the room temperature conductivity is in the range 1o-5 to 10-7(!Am)-‘. From the measurements of the conductivity versus time at 473 K and the XRD analysis, we explained that the increase of conductivity before 2.5 h of annealing is the result of a decrease of the unstable hcp phase and an increase of the number of connected crystallites, the decrease of conductivity after 2.5 h of annealing is due to the decrease of the number of connected crystallites. The decrease in the number of connected crystallites makes the intercrystalline potentie1 barrier higher, thus decreasing the conductivity. The reason for the increase of the activation energy is the decrease of defects between connected crystallites. These defects may introduce states in the energy map of C,, films. The deviation of the conductivity from straight lines in the high-temperature region in the 0 versus l/Tplots are due to unstable structural phases which transform toward a stable fee phase in the annealing process.

Acknowledgement

This work was supported Foundation of China.

by the National

Science

References [I] H. W. Kroto, J. R. Heath, S. C. O’Brian,

R. F. Curl and R. E. Smally, Nature, 318 (1985) 162. [2] W. Krltschmer, L. D. Lamb, K. Fostiropoulos and D. R. Huffman, Nature, 347 (I 990) 354. [3] M. S. Dresselhaus, G. Dresselhaus and P. C. Eklund, J. Mater. Res., 8(8) (1993) 2054. [4] A. F. Hebard, M. J. Rosseinsky, R. C. Haddon, D. W. Murphy, S. H. Glarum, T. T. M. Palstra, A. P. Ramirez and A. R. Kortan, Narurr, 350 (I 99 I ) 600. [5] M. J. Rosseinsky, A. P. Ramirez, S. H. Glarum, D. W. Murphy, R. C. Haddon, A. F. Hebbard, T. T. M. Plastra, A. R. Krotan, S. M. Zahurak and A. V. Makhija, Phys. Rev. Lets., 66 (1991) 2830. [6] S. P. Kelty, C. C. Chen and C. M. Lieber, Nature, 352 (1991) 223. [7] K. Tanigaki, T. W. Ebbesen, S. Saito, J. Mizuki, J. S. Tsai, Y. Kubo and S. Kuroshima, Narure, 352 (1991) 222. [8] J. Mort, K. Okumura, M. Machonkin, R. Ziolo, D. R. Huffman and M. I. Ferguson, Chem. Phys. Letr., 186(1991) 281. [9] J. Mort, R. Ziolo, M. Machonkin, D. R. Huffman and M. I. Ferguson, Chem. Phys. Left., 186 (1991) 284. [IO] J. Mort, M. Machonkin, R. Ziolo, D. R. Hutfman and M. I. Ferguson, Appi. Phys. Lert., 60 (1992) 1735. [I 11 R. C. Haddon, A. F. Hebard, M. J. Rosseinsky, D. W. Marphy, S. J. Duclos, K. B. Lyons, B. Miller, J. M. Rosamilia, R. M. Fleming, A. R. Kortan, S. H. Glarum, A. V. Makhija, A. J. Muller. R. H. Eick, S. M. Zahurak, R. Tycko, G. Dabbagh and F. A. Thiel, Nature, 350 (1991) 320. [I21 A. Hamed. Y. Y. Sun, Y. K. Tao, R. L. Meng and P. H. Hor, Phys. Rev., B, 47 (1993) 10873. [I31 H. Yonehara and C. Pat, Appl. Phys. Letr., 61 (1992) 575. [I41 C. Wen. J. Li, K. Kitazawa, T. Aida, I. Honma, H. Komiyama and K. Yamada, Appl. Phys. Lett., 61 (1992) 2162. [I51 R. K. Kremer, T. Rabenau, W. K. Maser, M. Kaiser, A. Simon, M. HaluSka and H. Kuzmany, Appl. Phys., A56 (1993) 211. 1161 T. Aral, Y. Mursakami, H. Suematsu, K. Kikucki, Y. Achiba and I. Ikemoto, Solid Stare Commun. 84 (1992) 827. [I71 D. Sarkar and N. J. Halas, Appl. Phys. Left., 63 (1993) 2438. [I81 P. J. Benning, J. L. Martins, J. H. Weaver, L. P. F. Chibante and R. E. Smalley, Science, 252 (1991) 1417. [I91 J. H. Weaver, J. L. Martins, T. Komedd, Y. Chen, T. K. Ohno, G. H. Kroll, N. Troullier, R. E. Ha&en and R. E. Smalley, Phys. Rev. Left., 66 (1991) 1741. [20] P. A. Heiney, J. E. Fisher, A. R. McGhie, W. J. Romanow, A. M. Denenstein, J. P. McCauley, Jr., A. B. Smith and D. E. Cox, Phys. Rev. Lerr., 66 (1991) 291 I. [2l] G. P. Kochanshi, A. F. Hebard, R. C. Haddon and A. T. Fiory, Science, 255 (I 992) 184.