ARTICLE IN PRESS

Science and Technology of Advanced Materials 7 (2006) 772779 www.elsevier.com/locate/stam

Electrical properties of ITO/benzylated cyclodextrins (b-CDs (Bz))/Al diode structures

M. Bouzitouna, C. Dridia,, R. Ben Chaabanea, H. Ben Ouadaa, H. Gamb, M. Majdoubb
a

partement de Physique, Faculte des Sciences de Monastir, 5019 Monastir, Tunisia Laboratoire de Physique et Chimie des Interfaces, De b `res et des Biopolyme `res, De partement de Chimie, Faculte des Sciences de Monastir, 5019 Monastir, Tunisia Laboratoire des polyme Received 31 May 2006; received in revised form 27 September 2006; accepted 2 November 2006 Available online 19 January 2007

Abstract Investigations of the electrical characteristics of benzylated cyclodextrins (b-CDs (Bz)) diodes are reported. We present currentvoltage characteristics and impedance spectroscopy measurements performed on partially benzylated cyclodextrins b-CDs (Bz) thin lms in sandwich structures ITO/b-CDs (Bz)/Al. The static electrical characterizations show a space charge limited conduction (SCLC) and a conductivity with power low frequency behavior characteristic of a hopping transport in disordered materials. The impedance spectra can be discussed in terms of an equivalent circuit model designed as a parallel resistor RP and capacitor C P network in series with resistor RS . We extract numerical values of these parameters by tting experimental data. Their evolution with bias voltages has shown that the SCLC mechanism is characterized by an exponential trap distribution. We estimated from the capacitance voltage characteristic an acceptor concentration of about 1011 cm3 due to trap states. r 2006 NIMS and Elsevier Ltd. All rights reserved.
Keywords: ITO/b-CDs (Bz)/Al diode; IV characteristics; Space charge limited current (SCLC); Impedance spectroscopy (IS)

1. Introduction Organic materials have recently attracted a great interest as semi conducting materials [13] and have been successfully applied to electroluminescent devices [3,4], eld-effect transistors [5,6], solar cells [7,8] and chemical sensors [9,10], in particular, molecular hostguest recognition systems such as calixarenes and cyclodextrins [1016]. The latter have good physical and chemical stability, compatibility with large eld applications and lead to low-cost devices. Among synthetic host molecules, cyclodextrins (CDs) have emerged as an ideal candidate for this pursuit due to its well-dened molecular cavity and its ability to accommodate a variety of guest molecules [17]. Nevertheless, the strong solubility of CDs in water makes their use difcult in the development of chemical sensors working in aqueous media. In our group, different immobilization methods
Corresponding author. Institut Superieur des Sciences Appliquees et de Technologie de Sousse, Cite Ettafala, 4003 Ibn Khaldoun Sousse, Tunisia. Tel.: +216 98 940 256; fax: +216 73 332 258. E-mail address: dridi2@yahoo.fr (C. Dridi).

have been used to overcome this problem: the b-CDs molecules were xed to silica insulator surfaces; either by chemically grafting polymethyl-hydrosiloxanes (PMHS) chains as a coupling agent [18] or by physically incorporating CDs in plasticized poly(vinyl chloride) (PVC) [19]. In previous work, we have chemically modied b-CDs (b-CDs (Bz)) and used them as sensitive membranes for heavy metal cations detection (such as Pb2 and Cd2 [20]. This has given us a great stimulus to the deeper study of their electrical and dielectric properties. Impedance spectroscopy (IS) has proved to be a powerful tool for studying the transport mechanisms and relaxation processes in organic devices [21,22]. The objective is to establish a correlation between chemical structure, in terms of number of benzyl moieties, and electrical performances toward an ion sensor with optimized sensitivity. Indeed, we present in another work [23] a comparison between CD grafted with 5 and 10 benzyl moieties in terms of heavy metal detection. In this paper, we have investigated the electrical properties of this partially benzylated b-cyclodextrin whose synthesis was presented in the previous work [20]. We investigated charge-carrier injection and transport

1468-6996/$ - see front matter r 2006 NIMS and Elsevier Ltd. All rights reserved. doi:10.1016/j.stam.2006.11.006

ARTICLE IN PRESS
M. Bouzitoun et al. / Science and Technology of Advanced Materials 7 (2006) 772779 773

Al -V +V -CD (Be) ITO Glass substrate

Fig. 1. ITO/b-CD(Bz)/Al diode heterostructure.

3. Results and discussion 3.1. Surface morphology The produced lms are smooth and homogeneous in surface topography as shown in the AFM image in Fig. 2. The surface shows a very low roughness (RMS) of $0:5 nm measured over a 500 500 nm2 area. The thickness was estimated by the AFM-scratching technique to be about 20 nm. 3.2. UVVis study

processes of b-CDs (Bz) diodes via IV characteristics as well as impedance spectroscopy. The diode structure studied is a metal/polymer/metal consisting of indium tin oxide as positive contact and an aluminum electrode as the negative contact as shown in Fig. 1. The IS measurements at different bias voltages have given a deeper insight into transport mechanism of the device. Moreover, the obtained results were tted to an electrical equivalent circuit in order to extract the relaxation time in these materials.

UVVis spectra of b-CDs (Bz) performed in solution and on thin lm samples, exhibit one signicant absorption band lmax 260 nm as shown in Fig. 3a. The estimated optical band gap is about 4.4 eV giving a device electronic structure shown in Fig. 3b. 3.3. Currentvoltage characteristics The measured IV characteristics show typical diode behavior with a threshold bias voltage of 2.5 V. As it can be seen in Fig. 4a, the current was observed in both forward and reverse bias voltage. The electrically symmetric IV characteristic, for low voltages, can be explained by the localized-state theory with defects providing the localized gap states [24]. Indeed, if the defects are amphoteric, it is possible to observe electrically symmetric behavior at low voltages. The asymmetric IV characteristic at higher voltages is attributed to the difference of injection barriers to electron and hole due to different work functions for the ITO anode (4.7 eV) and the Al cathode (4.3 eV). The IV characteristic in loglog plots of ITO/b-CDs (Bz)/Al is shown in Fig. 4b. The current dependence of applied voltage appears to follow power law behavior J / V m with two regimes corresponding to: (1) An ohmic region with J / V at low voltage. Indeed, for the case of existence of a single set of shallow traps in

2. Experimental details Thin lms were obtained by spin-coating b-CDs(Bz)/ chloroform solutions on ITO glass substrates which were precleaned by successive ultrasonic treatment for 20 min in acetone and methanol followed by drying with nitrogen gas (ITO-thickness 100 nm, sheet resistance 20 O=cm2 . Following the spin-coating process, the lms were annealed for 30 min in an oven at 80  C under ambient atmosphere. The indium tin oxide (ITO) thin lms have been used, due to their good efciency as a hole injecting material into organic thin lm. The aluminum top electrode was deposited on b-CDs (Bz) thin lm by thermal evaporation through appropriately shadow masks in high vacuum conditions 106 Torr. The currentvoltage characteristics of the ITO/b-CDs (Bz)/Al devices were measured from an applied bias of 55 V by using a Keithley 236 source measure unit. In general, the excitation potential for AC measurements is given by: V V 0 V mod cosot, (1)

with V 0 is the DC bias and V mod is the oscillation level and o=2p is the frequency. In our case, the measurements were performed in the following conditions V 0 : 023:5 V and V mod of 50 mV over a frequency range of 500 Hz13 MHz using a computer controlled HP 4192A LF. All electrical measurements were performed in dark and at room temperature. The UVVisible absorption spectra were recorded with a Perkin-Elmer UVVis spectrophotometer (Lambda 35) on b-CDs (Bz) thin lms spin-coated on spectrosil glass substrates to avoid the strong absorption of ITO in the UV.

Fig. 2. AFM topography of a spin-coated partially benzylated cyclodextrins (b-CDs (Bz), image at 3D.

ARTICLE IN PRESS
774 M. Bouzitoun et al. / Science and Technology of Advanced Materials 7 (2006) 772779

0.9

Solvent:Chloroform CD-Bz (thin film) CD-Bz (Solution)

3.0x10-6 2.0x10-6 1.0x10-6

Absorbance

I (A)

0.6

0.0 -1.0x10-6 -2.0x10-6

0.3

-3.0x10-6 -4.0x10-6 -5 -4 -3 -2 -1

0.0 200

300 (nm)

400
1x10-5

0 1 V (Volts)

LUMO Electrons Anode (ITO, 4.7 eV) EGAP

Log I (A)
10-6
B 2.0

Cathode (Al, 4.3 e V)

10-7
1.3

Holes HOMO -CDs (Bz),EGAP= 4.4 e V

Fig. 3. (a) UVVis spectra of b-CDs (Bz) and (b) energy levels diagram of ITO/b-CDs (Bz)/Al.
0.1 10-8
A

1 LogV (Volts)

solids, at low voltages, electrical conduction is Ohmic and the current density is described by [25] V , (2) d where q: is the electronic charge, m: the charge carrier mobility, n0 : the free carrier density, V: the applied voltage and d: the lm thickness. (2) A trapped charge limited region with J / V 2 . As the voltage increases, the current is controlled by space charge and follow the relation [25,26]: J qmn0 8 V2 eym 3 , (3) 9 d where e is the organic material permittivity, y relating the proportion of trapped charges nt ; pt to free charges n; p then y n=n nt p=p pt where n and nt are the density of free and trapped electron respectively, p and pt the density of free and trapped hole respectively. J These two regimes show clearly a space charge limited current (SCLC) process [27,28]. This is a similar behavior as observed in previous works for several ITO/organic lm/Al diode structures [2936]. The organic layers used were a

Fig. 4. (a) Currentvoltage characteristics of the ITO/b-CD (Bz)/Al and (b) loglog plot of the IV curve of Fig. 4a. The slopes of AB and BC lines indicate two regions ohmic and SCLC with exponents 1.22 and 2.04, respectively.

conjugated polymer (PPV), a small molecule Alq3 , an alternating block copolymer and other different macromolecules (diazoamino glyoxime (DAG) and anthylvinyl benzene (AVB)) [2936]. These reports have considered that this SCLC mechanism is attributed to a bulk-limited conduction process. However, de Boer and Morpurgo [37] have shown recently; that surface traps cause a large change in the electrostatic prole of the bulk material thereby profoundly affecting the behavior of SCLC. Their calculations illustrate how the combined effect of surface and bulk traps results in features in the IV curves that have been so far attributed to different physical mechanisms. The crucial point is that in high-purity samples, the total amount of surface traps can dominate over the total amount of bulk traps, even for sizable contact separations [37]. 3.4. Impedance spectroscopy 3.4.1. Go and Co measurements Impedance spectroscopy was applied to understand the dielectric behavior of the ITO/b-CDs (Bz)/Al device.

ARTICLE IN PRESS
M. Bouzitoun et al. / Science and Technology of Advanced Materials 7 (2006) 772779 775
0V 0.5V 1V 1.5V 2V 2.5V 3V 3.5V

The variation of conductance and capacitance, of b-CDs (Bz) derivative based diodes versus frequency at different bias voltages are plotted as shown in Figs. 5 and 6, respectively. The conductance characteristic remains constant at low frequencies and power-law behavior was observed at higher ones. Indeed, a hopping transport mechanism can be postulated. In general, the trend of the conductance with frequency, in disordered materials, obeys the following relation [38]: Go G dc G ac o where G ac o Aos , (4) where G dc is the dc conductance, o the angular frequency of the applied excitation, A constant, s the critical exponent 0oso1. As shown in Fig. 6 the capacitance decreases as the frequency increases (capacitance curves are off-set by 1 pF for clarity). This effect can be ascribed to traps inside the device [39] as observed also in polymer light emitting diodes by Blom and coworkers [40]. In our case, it appears that the capacitance C exhibits distinct frequency dependence:  At low frequencies C decreases and the frequency at which this decrease stops is about 10100 kHz depending on the applied bias. Indeed, since the device is space charge limited, the ac frequency modulates the number of charge carriers in the device. Nevertheless, the time scale for the built-up of charge carriers is given by the transit time tt of the injected carriers. At low frequencies the built-up of the space charge is fast enough to follow the ac modulation. Given that tt is nite the corresponding current lags behind the ac modulation excitation and this result in an additional inductive (negative) contribution to the capacitance.  For higher frequencies the space charge cannot be redistributed in a period of the applied voltage and the measured capacitance equals to the geometrical value, C 0 e0 er A=d.
1E-5

1E-8

Capacitance (F)

1E-9
Capacitance (F)
E1-11

1E-10

10

1E-11

100 1000 f (KHz)

10 000

1E-12

10

100 f (KHz)

1000

10 000

Fig. 6. Frequency dependent capacitance of the ITO/b-CD(Bz)5 /Al diode at different bias voltage.Th respective capacitance traces are off-set by 1 pF for clarity.

Moreover, the slow decrease of C and strong increase of G with frequency reect the relaxation of permanent dipoles present in the material. The disordered nature of the organic material gives rise to a distribution of dipolar relaxation times, (the evolution of the latter with bias voltage will be discussed in Section 3.4.3) and the dielectric response can be accurately described by the ColeCole representation [41]. 3.4.2. Colecole representation The IS technique consists in the measurement of the electrical impedance Z as a function of the frequency of the input signal over a wide frequency range. The collected data may be visualized as a Nyquist diagram or colecole plot, represented by the imaginary component Z 00 of the impedance as a function of Z 0 the real component. The ColeCole plots of ITO/b-CDs (Bz)/Al device at several dc bias voltages as seen in Fig. 7, display a single semicircle and a net decrease in diameter with increasing applied dc bias. The strong dependence of the impedance with the bias indicates the presence of a depletion layer in the sample with a resistance decreasing with increasing bias. Capacitance voltage measurements (see Section 3.5) conrm the presence of a depletion layer in forward bias before the injection become signicant. In this case two semicircles should be obtained on the plot of ImZ vs Re(Z), one representing junction and the other representing the bulk behaviour. In our case, only one semicircle is observed because the bulk and the junction capacitances are not sufciently different. Thus, the semicircle diameter corresponds to the sum of bulk and junction resistances. Analog results has been obtained for PPV [42] and AVB [36] based devices. A single semicircle in a Nyquist plot suggests an equivalent electrical circuit, shown in Fig. 8, composed of a single parallel resistor RP and a capacitor C P network placed in series with resistance RS [21]. We plotted the

1E-6
G (1)

1E-7

bias 0.0V s=0.97 bias 0.5V s=0.93 bias 1.0V s=0.89 bias 1.5V s=0.78 bias 2.0V s=0.67 bias 2.5V s=0.60 bias 3.0V s=0.54 bias 3.5V s=0.50

1E-8 0.1

10

100 f (kHz)

1000

10000

Fig. 5. Conductance plot as a function of the frequency in different bias voltages.

ARTICLE IN PRESS
776 M. Bouzitoun et al. / Science and Technology of Advanced Materials 7 (2006) 772779

1.2x107
Z "()

Bias 0.0V Bias 0.5V Bias 1.0V Bias 1.5V Bias 2.0V Bias 2.5V Bias 3.0V

106 0.54

Im Z()

105

6.0x106

0.0 0 1x107 Z' ()

Fig. 7. Colecole plots of the ITO/b-CD(bZ)/Al device at several bias voltage.

104 2x107 3x107 102 103 104 105 Re Z() 106

Fig. 9. The colecole plot of the ITO/b-CD(bZ)/Al device at 2.5 V bias voltage (in loglog representation). The slope of 0.5 means the curvature is semicircle.

Rp Rs

where o is the angular frequency of the ac excitation and o0 1=RP C P the proper angular frequency of the circuit. Eliminating the angular frequency we obtain the following equation relation:    RP 2 R2 ReZ RS ImZ2 P . 2 4 (6)

Cp ITO / -CD(bZ) -CD(bZ) bulk and -CD(bZ) /Al junction

Fig. 8. The electrical equivalent circuit of the ITO=bCDBz5 =Al diode which consist of a single parallel resistor RP and capacitor C P network with a series resistance RS .

data in loglog scale (Fig. 9). The slope of 0.5 means the curvature is a semicircle [21,41]. The obtained slopes are in the range of 0.460.68 for high and low bias voltages. It suggests a single characteristic relaxation time. We presented in Fig. 10(a) the real part of the impedance of ITO/b-CDs (Bz)/Al device as a function of frequency at 0, 1.5 and 3.5 bias voltages. The maximum Re(Z) value corresponds to the sum of the device resistance and the resistance of the ITO substrate. At higher frequencies, Re(Z) (Fig. 10a) and Im(Z) (Fig. 10b) are more frequencyindependent. 3.4.3. Simulation of impedance spectroscopy results The impedance of the equivalent circuit is given by " #   RP RP o=o0 Zo RS j 1 o=o0 1 o=o0 2 ReZj ImZ, 5

The minimum Re(Z) value observed at high frequencies corresponds to the RS value which is bias and frequency independent and should be attributed to the hole injecting interface ITO/b-CDs (Bz) which can be regarded as ohmic. In this device RS is about 289.5 O (for 0 V bias). The maximum Re(Z) value at low frequencies corresponds to the sum of RS and the b-CDs (Bz) resistance RP which is strongly bias dependent. Fig. 11 shows the measured real and imaginary parts of the impedance as a function of the frequency for the diode ITO/b-CDs (Bz)/Al at dc bias voltage of 0 V. Symbols are measured data and the solid lines are the tting results using the correspondent equivalent circuit of Fig. 8. The t of experimental results by Eq. (5) give the parameters RP and C P of the circuit which are summarized in Table 1. The interfacial capacitance is negligible, which suggests that there are negligible insulating barriers such as oxide or air gap formed between the b-CDs (Bz) thin lm and the top and/or bottom electrodes. The variation of the tting parameters of the ITO/b-CDs (Bz)/Al device are shown in Fig. 12. The RP decreases as the dc bias voltage increases while C P is almost independent of bias voltage. The independence of C P with applied bias is exactly the behavior expected for the SCLC mechanism, where the total charge in the device is equal to C P V .

ARTICLE IN PRESS
M. Bouzitoun et al. / Science and Technology of Advanced Materials 7 (2006) 772779 777 Table 1 Electrical parameters of the device calculated from the t of the experimental data Bias (V) 0 0.5 1 1.5 2 2.5 3 3.5 289.5 292.7 292.7 289.7 300.7 271.1 276.5 311.4 Rs (O) Rp (M O) 29.01 24.40 13.75 3.92 2.69 1.96 1.61 1.37 2.24 1.79 1.70 2.72 2.02 2.07 2.07 1.86 C p (pF) t (ms) 0.400 0.290 0.150 0.070 0.040 0.025 0.021 0.016

3.5x107 3.0x107 2.5x107

Z '()

2.0x107 1.5x107 1.0x107 5.0x106 0.0 1 10 100 f (KHz) 1000 10 000

0V 1.5V 3.5V

30 25 20

RP (M) CP (pF)

1.5x107

15 10

Ln RP

10

1.9

1.0x107
Z "()

1 Ln V

5 5.0x106
0V 1.5V 3.5V

0 -0.5 0.0 0.5 1.0 1.5 2.0 Bias (V) 2.5 3.0 3.5 4.0

0.0 1 10 100 f (KHz)

Fig. 10. Frequency dependence of (a) real part and (b) imaginary part of the impedance for different bias: 0, 1.5 and 3.5 V. Symbols are experimental data and the solid lines are the tting data using the equivalent circuit represented in Fig. 8.

1000

10 000

Fig. 12. The variation of the tting parameters of the ITO=bCDBz5 =Al diode using the equivalent circuit of Fig. 8. The inset shows the plot of log RP vs log V, and m is obtained as about 1.9 by the linear tting of the data.

From the SCLC with an exponential trap distribution theory, voltage dependent current for one carrier dominated transport, hole in this case, is given by [21] J KV m1 =d 2m1 , (7)

3.0x107 2.5x107
(Z', Z") ()

ITO / -CDs(Bz)/Al Bias = 0V

where d and K are the thickness of the lm and a constant, respectively. The voltage dependence of RP is thus described by RP aV =JaV m . (8) The plot of log RP vs log V is shown in the inset of Fig. 12. The linear behavior of log RP vs log V implies that conductance of the hole in b-CDs (Bz) thin lm is consistent with a SCLC mechanism with exponential trap distribution where the corresponding m is about 1.9. The carriers may be trapped by chemical impurities and/or structural defects. The range of relaxation times, for different bias voltage, is indicative of dipolar relaxation. 3.5. Capacitancevoltage characteristics More quantitative information about trap concentration in our device can be obtained from capacitancevoltage

2.0x107 1.5x107 1.0x107 5.0x106 0.0 1 10 100 f (KHz) 1000 10 000 Z' Z"

Fig. 11. An example of tting the impedance data of the ITO=bCDBz5 =Al diode using a single parallel resistor RP and capacitor C P with a series resistance RS . The data are well tted for over than four decades. Symbols are experimental data and the solid lines the tting ones.

ARTICLE IN PRESS
778 M. Bouzitoun et al. / Science and Technology of Advanced Materials 7 (2006) 772779

2.4x1023 2.2x1023 2.0x1023

1/C2 ( F -2)

75 KHz

1.8x1023 1.6x1023 1.4x1023 1.2x1023 1.0x1023 8.0x1022 0.0 0.5 1.0 1.5 2.0 Bias (Volts) 2.5 3.0

trap distribution and relaxation time can be extracted from the space charge limited current theory with traps and the impedance spectroscopy formalism. Moreover, the slow decrease of C and strong increase of G with frequency reect the relaxation of permanent dipoles present in the material. The disordered nature of the organic material gives rise to a distribution of dipolar relaxation times. From the plot of 1=C 2 vs bias voltage of ITO/b-CDs (Bz)/ Al diode we deduce the acceptor concentration N A . References
[1] J. Simon, J.J. Andre, Molecular Semiconductors: Photoelectrical Properties and Solar Cells, Springer, Berlin, 1985. [2] C. Dridi, A.H. Said, O. Ouerghi, M. Chikhi, A.P. Legrand, M. Gamoudi, J. Davenas, H. Maaref, Synth. Met. 115 (2000) 97. [3] L.S. Hung, C. Chen, Mater. Sci. Eng. R 39 (2002) 143. [4] X. Xu, G. Yu, Y. Liu, D. Zhu, Displays 27 (2006) 24. [5] G. Horowitz, Adv. Mater. 10 (1998) 365. [6] R.B. Chaabane, A. Ltaief, C. Dridi, H. Rahmouni, A. Bouazizi, H.B. Ouada, Thin Solid Films 427 (2003) 371. [7] C. Brabec, V. Dyakonov, J. Parisi, N.S. Sariciftci, Organic Photovoltaics: Concepts and Realization, Springer, Berlin, 2003. [8] A. Ltaief, J. Davenas, A. Bouazizi, R.B. Chaabene, P. Alcouffe, H.B. Ouada, Mater. Sci. Eng. C 25 (2005) 67. [9] R. Mlika, H.B. Ouada, N.J. Renault, I. Dumazet, R. Lamartine, M. Gamoudi, G. Guillaud, Sensors and Actuators B 47 (1998) 43. [10] G. Harsenyi, Sensor Rev. 20 (2000) 98. [11] R. Mlika, H.B. Ouada, R.B. Chaabane, M. Gamoudi, G. Guilaud, N.J. Renault, R. Lamartine, Electrochim. Acta 43 (1998) 841. [12] R. Mlika, M. Gamoudi, G. Guillaud, M. Charbonnier, M. Romand, J. Davenas, N.J. Renault, R. Lamartine, A. Touhami, Mater. Sci. Eng. C 11 (2000) 129. [13] M.B. Ali, R.B. Chabanne, F. Vocanson, C. Dridi, N. Jaffrezic, R. Lamartine, Thin Solid Films 495 (2006) 368. [14] M.B. Ali, Ph.D. Thesis, Ecole Centrale de Lyon, 2000. [15] E.V. Dienst, W.I.I. Bakker, J.F.J. Engbersen, W. Verboom, D.N. Reinhoudt, Pure Appl. Chem. 65 (1993) 387. [16] M.B. Ali, R. Kalfat, H. Shi, J.M. Chovelon, H.B. Ouada, N.J. Renault, Sensors and Actuators B 62 (2000) 233. [17] S. Busse, M. DePaoli, G. Wenz, S. Mittler, Sensors and Actuators B 80 (2001) 116. [18] R. Kalfat, M.B. Ali, R. Mlika, F.F. Romdhane, N.J. Renault, Int. J. Inorg. Mater. 2 (2000) 225. [19] M.B. Ali, R. Kalfat, H. Shi, H.B. Ouada, J.M. Chovelon, N.J. Renault, Mater. Sci. Eng. C 6 (1998) 53. [20] M. Bouzitoun, R. Mlika, H. Gam, H.B. Ouada, M. Majdoub, H. Shi, Mater. Sci. Eng. C 26 (2006) 481. [21] S.H. Kim, K.H. Choi, H.M. Lee, D.H. Hwang, L.M. Do, H.Y. Chu, T. Zyung, J. Appl. Phys. 87 (2000) 882. [22] H.C.F. Martens, H.B. Brom, P.W.M. Blom, Phys Rev. B 60 (1999) R 8489. [23] M. Bouzitoun, C. Dridi, R. Mlika, R.B. Chaabane, H.B. Ouada, Symposium G, European-Materials Research Society, Nice, 2006. [24] S. Jeglinski, M. Hollier, J. Gold, Z.V. Vardeny, Y. Ding, T. Barton, J. Mol. Cryst. Liq. Cryst. 256 (1993) 555. [25] K.C. Kao, W. Hwang, Electrical Transport in Solids, Pergamon Press, Oxford, 1981. [26] F. Michelotti, F. Borghese, M. Bertolotti, E. Cianci, V. Foglietti, Synth. Met. 111112 (2000) 105. [27] M. Pope, C. Swenberg, Electronic Processes in Organic crystals, Oxford University Press, New York, 1982. [28] W. Bruetting, S. Berleb, A.G. Mueckl, Synth. Met. 122 (2001) 99. [29] P.W.M. Blom, M.J.M. de Jong, J.J.M. Vleggaar, Appl. Phys. Lett. 68 (1996) 3308.

Fig. 13. 1=C 2 vs bias voltage of ITO/b-CDs (Bz)/Al diode (the inset shows the CV characteristic).

spectroscopy. We show in the inset of Fig. 13 the bias voltage dependent capacitance which indicate the fully depleted b-CDs (Bz) thin lm. We observe from the evolution of 1=C 2 versus bias voltage (Fig. 13), a locally linear behavior between 1 and 3 V which conrm the presence of surface traps acting as acceptor in b-CDs (Bz) and allows as to estimate their concentration N A . Therefore, from the curve of 1=C 2 vs V for an abrupt junction, assuming that the traps are uniformly distributed in space, the trap concentration N A can be calculated using the following relation [42]: 1 2V D V 2 , 2 C A qe0 eN A (9)

where V D : the diffusion voltage and N A : the ionized acceptor concentration. The ionized acceptor concentration N A have been estimated to be 4:9 1011 cm3 and the diffusion voltage V D equal to 4.3 V. Indeed, involving an exponential trap distribution is remarkably in good agreement with the experimental observation listed above. However, this model cannot be attributed only to bulk limited conduction in b-CDs (Bz) active layer but trap states should be considered to explain C 2 vs V local linear behavior. 4. Conclusion In this paper, currentvoltage and electric impedance measurements were used to investigate the transport characteristics of holes in ITO/b-CDs (Bz)/Al diode structures. The fabrication and testing of a polymer diode has been reported and discussed. The device exhibited symmetric IV characteristics in both forward and reverse bias. The conduction of holes in a lm of b-CDs (Bz) is governed by space charge limited current characterized by a presence of exponential trap distribution. From ts of frequency dependent experimental data, the exponential

ARTICLE IN PRESS
M. Bouzitoun et al. / Science and Technology of Advanced Materials 7 (2006) 772779 [30] A.J. Campbell, M.S. Weaver, D.G. Lidzey, D.D.C. Bradley, J. Appl. Phys. 84 (1998) 6737. [31] U. Wolf, S. Barth, H. Bassler, Appl. Phys. Lett. 75 (1999) 2035. [32] D. Ma, I.A. Hummelgen, X. Jing, Z. Hong, L. Wang, X. Zhao, F. Wang, F.E. Karasz, J. Appl. Phys. 87 (2000) 312. [33] W. Brutting, S. Berleb, A.G. Muckl, Synth. Met. 122 (2001) 99. [34] G.D. Sharma, D. Saxena, M.S. Roy, Synth. Met. 123 (2001) 189. [35] S.H. Kim, J.W. Jang, K.W. Lee, C.E. Lee, S.W. Kim, Solid State Commun. 128 (2003) 143. 779 [36] A. Rihani, N. Boutabba, L. Hassine, S. Romdhane, H. Bouchriha, Synth. Met. 145 (2004) 129. [37] R.W.I. de Boer, A.F. Morpurgo, Phys. Rev. B 72 (2005) 73207. [38] A.K. Jonsher, Thin Solid Films 1 (1967) 213. [39] H. Bassler, Phys. Stat. Solidi B 175 (1993) 15. [40] W. Blom, C. Martens, J.N. Huiberts, Synth. Met. 121 (2001) 1621. [41] A.K. Jonscher, Dielectric Relaxation in Solids, Chelsea Dielectrics Press, London, 1983, p. 85. [42] M. Meier, S. Karg, M. Riess, J. Appl. Phys. 82 (1997) 1961.