Si

international journal of hydrogen energy 34 (2009) 5208–5212

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/he

Electrical and photovoltaic properties of Cr/Si Schottky diodes Beyhan Tatar*, A. Evrim Bulgurcuog˘lu, Pınar Go¨kdemir, Pelin Aydog˘an, Deneb Yılmazer, Orhan O¨zdemir, Kubilay Kutlu Faculty of Arts and Sciences, Department of Physics, Yıldız Technical University, Davutpas x a 34220, Istanbul, Turkiye

article info

abstract

Article history:

The electrical properties of the Cr/p-Si(111) and Cr/n-Si(100) junctions were investigated

Received 5 September 2008

through capacitance–voltage and current–voltage measurements, performed under dark

Received in revised form

and light conditions at room temperature. Diode parameters of Cr/Si Schottky diode like

15 October 2008

ideality factor and barrier height were obtained and variations of them were monitored as

Accepted 16 October 2008

a function of temperatures. Also, an attempt to explore the governing current flow

Available online 21 November 2008

mechanism was tried. The reverse biased I–V measurement under illumination exhibited anomalous behavior as well as high photosensitivity. The former was explained in terms of

Keywords:

minority carrier injection phenomenon. The photovoltaic parameters, such as open circuit

Cr/Si Schottky diodes

voltage and short circuit current were obtained as 370 mV and Isc ¼ 44.5 mA, respectively.

Electrical properties

ª 2008 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights

Photovoltaic properties

1.

Introduction

Silicon-based photovoltaic devices are preferable not only having mutual properties with the microelectronic technology but also having long lifetime (w20 years) and high efficiency [1]. Two kinds of photovoltaic devices, differing from each other through production method, exist today. One of them is in the form of Schottky (metal/semiconductor h M–S) junction whereas another one is in the form of either in homojunction (p–p and/or n–n) or in heterojunction (p–n and/or n–p). Moreover the latter can be made via the same semiconductor by different type doping, yielding relatively more complex structure compared to M–S one [2,3]. On the other hand, M–S structures have been studied extensively because of the very importance and critical components in integrated circuit technology. Moreover, it is a very attractive tool for the characterization of semiconductor materials [4–8]. The electrical properties of the Schottky contact depend on the density of interface states which play crucial role on determination of

reserved.

diode parameters such as Schottky barrier height (F) and ideality factor (n). Besides, the control of interface property is very promising for device performance, stability and reliability [9–11]. In this frame, to investigate the behavior of electrical properties under light apart from dark condition, a thin metal film has to be coated on a semiconductor. Therefore, Schottky junction solar cells could be formed and tiny metal would allow to transmit the light in the semiconductor substrate. Hence, under illumination, photons having energy greater than energy band gap of semiconductor (hn > Eg) can generate electron-hole pairs in the depletion region of the semiconductor. Due to existing of electric field (i.e., built in field), the holes swept to the back contact while electrons gathered at the junction. Consequently, there would be an additional photocurrent over the usual dark diode current. Due to the thermionic emission of the holes that pass from the metal to the semiconductor, diode’s dark current is alike with the diode’s photocurrent under forward biased condition.

* Corresponding author. Tel.: þ90 212 383 42 83; fax: þ90 212 383 42 06. E-mail address: [email protected] (B. Tatar). 0360-3199/$ – see front matter ª 2008 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijhydene.2008.10.040


5209

Contrary to that, under reverse biased condition, the dark current in a Schottky barrier diode is several times higher than the one in the p–n junction, so that the open circuit voltage (Voc) of a Schottky solar cell would be lower than the p–n junction solar cell; yielding poorer efficiency compared to p–n solar cells. On the other hand, the main advantage of Schottky barrier solar cells is easiness of production such as they do not require high temperature processing, and thus, the processing cost is reduced remarkably [12,13]. This work is relevant with the research activities on electrical properties of silicon-based devices for production of photovoltaic electricity. The photovoltaic electricity can be used for hydrogen generation which is a clean renewable fuel through water electrolysis [14–17]. The photovoltaic properties of Cr/Si Schottky junctions and their fundamental diode parameters were investigated by using current–voltage (I–V) measurements at room temperature. Additionally, diode parameters such as n, F were monitored as a function of studied temperature interval (300–380 K).

2.

Experimental

The Cr/Si Schottky barrier diodes were prepared using nSi(100) and p-Si(111) substrates, whose resistivities are 5–10 Ucm and 11–20 U-cm, respectively. Before deposition, substrates were cleaned by standard RCA cleaning procedure, in which organic and inorganic contaminants were removed by H2SO4 þ H2O2 þ H2O (2:1:1 volume) and HCl þ H2O2 þ H2O (1:1:2 volume) solutions, respectively. As a last step before metal coating, Si substrates were immersed into a 5% HF solution. Subsequently, Cr metal which has 99.9% purity was evaporated on cleaned surface of n-type and p-type silicon substrates by e-beam evaporation method at room temperature under w105 Torr pressure. The produced film thickness was in situ measured around 2 nm through thickness profiler, equipped to the e-beam system at hand. For investigating electrical properties of Cr/Si Schottky diodes, InGa and Ag metals were used for n-type and p-type semiconductor as ohmic back contacts, respectively. InGa was front contact for both MS structures. The photovoltaic properties of junctions were investigated by I–V measurements, performed under room temperature in dark/light conditions and in the range between 250 K and 375 K in evacuated cryostat (103 Torr) in dark condition. The incident power density of light illumination was 100 W/m2 provided by the tungsten lamp. Keep in mind that the measurements were performed with computer controlled Keithley 6517 multimeter. The capacitance–voltage (C–V) measurements were performed at various frequencies by the use of HP 4192A impedance analyzer.

3.

Results and discussions

Electrical characteristics of both Cr/n-Si and Cr/p-Si Schottky diodes were monitored as a function of forward and reverse bias conditions at studied temperature interval (300–380 K) and illustrated in Fig. 1. As depicted in Fig. 1, Cr/n-Si and Cr/pSi Schottky diodes have fairly well rectifying property (w103) and became deteriorate as ambient temperature increases.

Fig. 1 – Semi-logarithmic I–V curves as a function of temperature for (a) Cr/n-Si and (b) Cr/p-Si Schottky diodes.

For Schottky diodes, provided that the effect of minority carriers on the V > 3kT/q region are ignored, thermionic emission (TE) model is valid. This theory suggests that the following I–V relationship is hold qV 1 (1) I ¼ I0 exp hkT where V is the applied voltage, k is the Boltzman constant, T is the temperature, h is the ideality factor and I0 is the reverse saturation current and given as; qFB (2) I0 ¼ AA T2 exp kT with A the diode area, A* the Richardson constant, q electron charge and FB the barrier height. Ideality factor of Cr/Si Schottky diodes was calculated from the slope of the linear

5210


part of the forward bias semi-logarithmic I–V characteristic and can be expressed as q dV (3) n¼ kT dlnI Moreover, FB can be extracted through determination of I0 which was the extrapolated value of forward lnI–V curve. Therefore, FB could be estimated through the following equation kT lnAA T2 (4) FB ¼ I0 q Fig. 1a and b showed the temperature dependence of I–V characteristic of the Cr/Si junctions. The ideality factor and barrier heights were calculated as mentioned above and demonstrated in Table 1. Temperature variations of ideality factor and barrier height of Cr/n-Si and Cr/p-Si are depicted in Figs. 2 and 3, respectively. Obviously, the ideality factor decreases as the temperature increases and the barrier height increases as the temperature increases. This type of dependency on temperature can be figured out due to the inhomogeneous of barrier height and/or the interfacial layer of the metal–semiconductor [18]. As shown in Table 1, the barrier heights of Cr/n-Si and Cr/ p-Si changed between 0.8 eV and 1.1 eV while the ideality factor varied between 1.6 and 1.1. Case of h greater than unity (h > 1) implies Schottky diode is not ideal [3,18]. One reason for that might be the density of the interface state exists between metal and semiconductor junction. Schottky emission, Poole-Frenkel, space charge limited (SCLC), etc. are the possible conduction mechanisms in the studied junctions [3]. The governing conduction mechanism in Cr/Si Schottky diodes at hand could be figured out from the power of lnI–lnV curve. Power of the curve greater than 2 (m > 2) indicates SCLC mechanism whereas being equal to 1 (m ¼ 1) imply ohmic character. When the power lies between 1 and 2 impose either Schottky or Poole-Frenkel conduction mechanism. In the light of information above, let us analyses the governing mechanism on structures at hand. As shown in Fig. 4, two linear regions are eventual in lnI–lnV curve of Cr/n-Si junction. One of them is in the lower electric field region (0.1 < VF < 0.5) where the power is in the proximity of unity (ohmic mechanism) and the other is in the higher electric field region (0.5 < VF), where the power is equal to 6.41 (SCLC mechanism). On the other hand, similar analysis is carried out for Cr/p-Si Schottky junction. In the lower electric field region where the power is equal to 1.5, the conduction mechanism

Fig. 2 – Variation of ideality factor (n) with temperature on Cr/n-Si and Cr/p-Si Schottky diodes.

might be through Schottky emission and in the higher electric field region, where the power is equal to 3.3, the current flows by SCLC mechanism. These results indicate that in the higher electric field region, SCLC is the common current flow mechanism for both Cr/n-Si and Cr/p-Si Schottky diodes. However, in the low electric field regions, two distinct mechanisms are eventual for MS junctions at hand; ohmic for Cr/n-Si and Schottky and Poole-Frenkel for the other. For Schottky and Poole-Frenkel conduction mechanisms, the current equation should be reconsidered as ! 1 bE2 FB (5) I ¼ I0 exp kT

Table 1 – The diode parameters of Cr/n-Si and Cr/p-Si Schottky diodes. Temperature (K) 300 320 340 360 380

Cr/n-type Si

Cr/p-type Si

h

fbn (eV)

h

fbn (eV)

1.60 1.51 1.41 1.27 1.11

0.85 0.88 0.93 0.97 1.04

1.58 1.56 1.52 1.50 1.46

0.81 0.87 0.90 0.95 1.02

Fig. 3 – Variation of barrier height (FB) as a function of temperature on Cr/n-Si and Cr/p-Si Schottky diodes.

5211


Fig. 4 – lnI–lnV characteristic of Cr/n-Si and Cr/p-Si Schottky diodes.

where E is the electric field and b is a constant and given as 12 e Q (6) b¼ b 330 with b is constant and equal to 4 for Schottky mechanism and unity for Poole-Frenkel mechanism, respectively. Theoretical relationship of 2 bs ¼ bpf(e3/P330)1/2 is hold for the specified conduction mechanisms [3]. Here, e is the dielectric constant of the natural SiO2 insulator layer on the semiconductor, e0 is the dielectric permittivity of free space. Theoretical bs and bpf are calculated and their values are obtained as 3.11 105 eV m1/2 V1/2 and 6.22 105 eV m1/2 V1/2, respectively. The experimental value of b (obtained as 3.6 105 eV m1/2 V1/2 by fitting the lower field electric region of lnI–V1/2 characteristic at low electric field side) is close to theoretical bs value and hence the mechanism is probably Schottky type. Fig. 5a and b showed the current–voltage values of Cr/n-Si and Cr/p-Si Schottky junction which were measured in dark and light condition under 100 mW light source. As clearly seen, the two diodes had photosensitive. Therefore, the photovoltaic parameters of the diodes (Voc ¼ the open circuit voltage, Isc ¼ short circuit current) were obtained and tabulated in Table 2. Photosensitivity of Cr/n-Si and Cr/p-Si was around 10 and 103 times, respectively. Hence, because of high photovoltaic properties, Cr/p-Si Schottky diodes could be used as a photodiode. If the photovoltaic parameters could be made even better, the diode could be a good candidate for solar cell technology. From now on, let us speculate the discrepancy of photosensitivity between the Schottky diodes at hand, especially reverse bias I–V behavior of Cr/p-Si under light. Generally, c-Si Schottky contacts are believed as majority carrier devices and minority carrier injection is disregarded. Though injection of minority carriers (electrons for p-type Si and holes for n-type Si) is usually ignored, one should have to

Fig. 5 – lnI–V characteristic of (a) Cr/p-Si and (b) Cr/n-Si Schottky diodes, measured at room temperature in dark and light conditions.

consider the phenomenon especially at high forward bias condition where minority carrier injection and charge storage become significant. In other words, both electrons and holes transportation take place simultaneously. To do so, barrier height for minority carriers should be low compared to that of majority carriers. Indeed, the barrier height for minorities will

Table 2 – The photovoltaic parameters of Cr/n-Si and Cr/ p-Si Schottky diodes. Sample type Cr/n-Si Cr/p-Si

Voc (mV)

Isc (mA)

Photosensitivity (Ilight/Idark)

366 370

25.2 44.5

10 103

5212


photovoltaic properties were investigated through I–V measurements. The ideality factor of Cr/Si Schottky diodes decreased with the increase in temperature but the barrier height increased. Two different conduction mechanisms were identified for two types of diodes. Photovoltaic measurement indicated that Cr/p-Si Schottky diodes had 103 times sensitive to the light, proposing a good candidate as a photodiode. Speculation was attempted to resolve this issue.

references

Fig. 6 – The C–V characteristics of the Cr/p-Si Schottky diode at room temperature under various measuring frequencies (1 kHz–1 MHz).

decrease as the barrier height for majority carriers increase. Consequently, minority carrier injection occurs inevitably and produced an inversion layer at the interface where minority carrier amount exceeds the majority carrier amount. Those charges would not contribute to the surveying techniques unless certain conditions are fulfilled. One of them is as follows in dark condition: those stored minorities (electrons) will diffuse into neutral region of p-Si semiconductor under forward bias. In the mean time, additional holes should enter the neutral region of Schottky diode to preserve the charge neutrality, leading to increase in bulk conductivity [3,18]. This phenomenon is known as conductivity modulation. On the other hand, once I–V measurement is performed under light, additional electrons would be supplied over the one created due to illumination (i.e., second condition). Thus, the reverse bias current exceeds the forward biased current values in amount under light condition. The offered scenario seems more plausible when the capacitance–voltage (C–V) characteristic is reported (see Fig. 6). Comparing the C–V curve with the I–V curve measured at room temperature under dark condition, the change in capacitance (i.e., increase in capacitance value and the corresponding bias associated with the drastic increase and decrease in capacitance values) is due to the majority and minority carrier injection from electrodes into c-Si semiconductor [3].

4.

Conclusion

Chromium (Cr) metal was evaporated on n-Si(100) and p-Si(111) substrates at room temperature and then Cr/n-Si and Cr/p-Si Schottky diodes were produced. The electrical and

[1] Makita Y. Materials availability for thin film solar cells. In: Proceeding of future generation photovoltaic technologies: first NREL conference CP404; 1997. p. 704. [2] Markvart T. Solar electricity. Baffins Lane, Chichester, West Sussex PO19 1 UD, England: John Wiley & Sons; 2000. [3] Sze SM, Ng Kwok K. Physics of semiconductor devices. Hoboken, New Jersey: John Wiley & Sons, Inc.; 2007. [4] Cowley AM, Sze SM. Surface states and barrier height of metal–semiconductor systems. Journal of Applied Physics 1965;36:3212. [5] Nielsen OM. Influence of semiconductor barrier tunneling on the current–voltage characteristics of tunnel metal–oxide– semiconductor diodes. Journal of Applied Physics 1983; 54(10):5880. [6] Card HC, Rhoderick EH. Studies of tunnel MOS diodes I. Interface effects in silicon Schottky diodes. Journal of Physics D: Applied Physics 1971;4:1589. [7] Tatarog˘lu A, Altındal S x . The C–V–f and G/w–V–f characteristics of Al/SiO2/p-Si (MIS) structures. Microelectronic Engineering 2006;83:582–8. [8] Karatasx Sx . Comparison of electrical parameters of Zn/p-Si and Sn/p-Si Schottky barrier diodes. Solid State Communications 2005;135:500–4. [9] Altındal Sx , Karadeniz S, Tug˘luog˘lu N, Tatarog˘lu A. The role of interface states and series resistance on the I–V and C–V characteristics in Al/SnO2/p-Si Schottky diodes. Solid State Electronics 2003;47:1847. [10] Rhoderick EH, Williams RH. Metal–semiconductor contacts. 2nd ed. Oxford: Clarendon Press; 1978. [11] Chattopadhyay P, Daw AN. On the current transport mechanism in a metal–insulator–semiconductor (MIS) diode. Solid State Electronics 1986;29(5):555. [12] Kasap SO. Optoelectronics and photonics: principles and practices. New Jersey: Prentice-Hall Inc.; 2001. [13] Tyagi MS. Introduction to semiconductor materials and devices. Canada: John Wiley & Sons; 1991. [14] Kumari S, Chaudhary YS, Agnihotry SA, Tripathi C, Verma A, Chauhan D, et al. A photoelectrochemical study of nanostructured Cd-doped titanium oxide. International Journal of Hydrogen Energy 2007;32:1299–302. [15] Yamaguchi K, Udono H. Novel photosensitive materials for hydrogen generation through photovoltaic electricity. International Journal of Hydrogen Energy 2007;32:2726–9. [16] Buitrago RH, Risso GA, Cutrera M, Battioni M, Bernardez LD, Schmidt JA, et al. Polycrystalline silicon thin film solar cells prepared by PECVD–SPC. International Journal of Hydrogen Energy 2008;33:3522–5. [17] Sebastian PJ, Mathews NR, Mathew X, Pattabi M, Turner J. Photoelectrochemical characterization of SiC. International Journal of Hydrogen Energy 2001;26:123–5. [18] Kanicki J. Amorphous and microcrystal semiconductor devices II. London: Artech House; 1992.