based organic devices

(1)

Role of growth temperature on the frequency response characteristics of pentacene-

based organic devices

Yayun Shao¹, Yang Zhang¹, Wenqiang He¹, Chuan Liu², Takeo Minari³, Sujuan Wu¹, Min Zeng¹, Zhang Zhang¹, Xingsen Gao¹, Xubing Lu¹and J-M Liu^1,4

1Institute for Advanced Materials and Laboratory of Quantum Engineering and Quantum Materials, South China Normal University, Guangzhou, 510006, People’s Republic of China

2State Key Laboratory of Optoelectronic Materials and Technologies, School of Physics and Engineering, Sun Yat-Sen University, Guangzhou 510274, People’s Republic of China

3International Center for Materials Nanoarchitectonics, NIMS, Tsukuba, Ibaraki 305-0044, Japan

4Laboratory of Solid State Microstructures and Innovation Center of Advanced Microstructures, Nanjing University, Nanjing 210093, People’s Republic of China

E-mail:[email protected]@nju.edu.cn Received 6 October 2014, revised 10 December 2014 Accepted for publication 15 December 2014

Published 8 January 2015 Abstract

The ac frequency response characteristics (FRC) of organic thinfilm transistors and metal- insulator semiconductor diodes were highly improved by controlling the morphology and electrical characteristics of semiconducting pentacenefilms. The devices withfilms grown at 50 °C show much higher cutoff frequency and better frequency stability offlat-band voltage, as compared to those withfilms grown at other temperatures below or above. The improvement mainly originates from the maximumfield effect carrier mobility of 0.78 cm²V⁻¹s⁻¹and a small metal/organic contact resistance (R_c) obtained in the optimum thinfilm transistors. Our results indicate growth temperature precisely tunes thefilm microstructure and metal/semiconductor interface, which together determine the FRC of pentacene-based organic devices.

Keywords: organic thinﬁlm transistor, frequency response, pentacene (Someﬁgures may appear in colour only in the online journal)

1. Introduction

Organic thinﬁlm transistors (OTFTs) have been extensively studied in recent years due to their advantages of low cost, ﬂexibility, wide application, etc [1–4]. Limited by factors such as low carrier mobility and high metal/semiconductor contact resistance, the frequency response characteristics (FRC) of organic devices still cannot compete with those of inorganic devices. Recently, various efforts have been made to study the frequency response of organic semiconductor based devices through the viewpoints of complex impedance [5, 6], capacitance [7–10], output voltage [11, 12] and dis- placement current [13]. Luet alalso reported the air exposure and temperature effects on the RFC of pentacene-based

organic thin ﬁlm transistors [14, 15]. Various mechanisms such as channel length [7, 16], parasitic capacitance [17], charge trapping [14] and contact resistance [16] have been proposed to affect the FRC of organic devices. To date, the reported FRC is still far from the requirements for practical technology applications in high-speed devices such as radio frequency identiﬁcation (RFID) tags [13]. Therefore it is urgent for researchers to further clarify the mechanisms affecting their FRC and develop new ideas to improve frequency response behaviors.

The key parameter for frequency response is the cutoff frequency (f_c), which is critical for the normal operation of high speed OTFTs such as RFID tags [13]. Based on theoretical analysis [10,18], the cutoff frequency of the OTFT can

Semicond. Sci. Technol.30(2015) 035005 (7pp) doi:10.1088/0268-1242/30/3/035005

(2)

be deﬁned as f_c=^μ

π

− ,

V V L

( )

2

G T

2 where μ is the channel carrier mobility,V_Gis the gate voltage,V_Tis the threshold voltage, and L is the channel length. According to this expression, shortening of the channel lengthLwill be an effective way to increase thef_cvalue. However,Lis limited by the commonly used shadow mask technology, while photolithography technology is not a good choice for organic devices to define the source/drain area. Therefore, it is difficult to enhancef_c value by further decreasing L. Alternatively, increase of the carrier mobility and decrease of the threshold voltage will be feasible ways to further increasef_cof organic devices. In this work, we successfully tuned the carrier mobility and the threshold voltage by changing the substrate temperatures during the pentacene vacuum evaporation processes. Thefilm microstructure and surface morphology can be well controlled by the growth temperature. It was found that the surface morphology as well as the metal/organic interface plays a critical role on the FRC of pentacene-based organic devices.

2. Experimental details

Typical bottom-gate pentacene based MIS and OTFTs with top-contact electrodes were fabricated using the following conditions. A highly doped Si (100) wafer with a 100 nm thick thermally oxidized SiO₂layer was used as a substrate, where the Si wafer is used as the gate electrode and the SiO₂ layer acts as a gate insulator. The surface was ultrasonically cleaned in acetone and 2-propanol, followed by UV-O₃ treatment for 5 min. To improve the interfacial quality between gate insulator and organic semiconductor layer, one layer of poly (α-methylstyrene) (PS) with ∼30 nm thickness was deposited by spin-coating methods. The PS layer was annealed in air for 5 min at 120 °C. The 40 nm thick pentacene (Aldrich, purified with temperature gradient sublimation) film was deposited in a vacuum chamber onto the PS covered SiO₂ surface to form the organic semiconductor layer. To obtain different grain size and surface morphology, pentacene films were deposited at different substrate temperature with the same growth rate of 0.1 Å s⁻¹. Copper (Cu) electrodes were deposited by thermal evaporation through a metal mask to form the MIS and OTFT devices. Comb-shaped OTFTs with a channel length of 100μm were fabricated for capacitance–frequency (C–f) measurements. The detailed structures and measurement circuit of these devices can be found in our previous work [10,14,15]. The crystal structures and surface morphologies of the pentacenefilms were investigated by x- ray diffraction (XRD) (Philips X’pert MRD meter) and atomic force microscope (AFM) (Asylum Research) working in tapping mode. The average grain sizes of the pentacene films grown at different temperatures were estimated by using the image processing software of Image-Pro Plus. The dynamic ac characteristics of the MIS capacitor were measured by an LCR meter (Agilent E4980A) and the static dc characteristics of the transistors were measured by an Agilent B1500A semiconductor device analyzer. All the electrical

measurements were performed in dark under high vacuum (<5 × 10⁻³Pa) in a Jannis temperature variable probe station.

3. Results and discussions

Figure 1(a) shows the cross section structure of the OTFTs and MIS diodes for the ac frequency response and dc measurements. The C–fmeasurements were carried out to characterize the cutoff frequency of the pentacene-based OTFTs.

Figure 1(b) shows the C–f characteristics of comb-shaped OTFTs at typical growth temperatures of room temperature (RT), 50 °C and 70 °C. The biased gate voltage is−40 V. The fc values were estimated as the intersection points of the extrapolated ﬁtting lines and the capacitance of the low frequency region. C–f measurements were also carried out at various biased gate voltages from −10 V to −40 V, and the extractedf_cvalues were plotted as a function of gate voltages.

As shown in figure 1(c), thef_c values change with gate voltages as expected from theoretical prediction [10,18]. Under the same gate voltage, the OTFT with 50 °C grown pentacene film exhibits much higher f_c value compared to that of the other two samples. The 50 °C grown sample has anf_cvalue of 42.9 kHz at the gate voltage of −40 V, while the observed values are only 25.4 kHz and 23.4 kHz for RT and 80 °C samples. The mechanisms responsible for the significant differences of the cutoff frequency will be investigated in the following discussions.

To understand the FRC of the above devices, the surface morphology of the pentacene films deposited at different substrate temperatures was first studied by AFM measurements (figure 2(a)). The film grown at RT exhibits typical arborization structure and the grain size increases with the substrate temperature until 60 °C. With further increasing the substrate temperature to 70 °C and 80 °C, the grain morphologies change from arborization-like structure to mud- like structure, exhibiting significantly decreased uniformity in grain size distribution. For the 80 °C grownfilm, we can see clearly many small grains at the centre, and the overall grain sizes were greatly decreased. Figure2(b) illustrates the strong dependence of the grain sizes on the growth temperature of pentacenefilms. Accordingly, a moderate growth temperature around 50∼60 °C should be favorable to form well-ordered crystalline structure and large grain size.

The above characterization was associated with XRD analysis in a symmetric reflection (coupled θ–2θ mode, figure2(c)). Only (00l) reflection peaks were observed for all the films, signifying highly oriented pentacene films. The first-order (001) diffraction peak generally appears at 5.56°

and the corresponding d₀₀₁-spacing is ∼15.88 Å. Thus, our pentacene films follow the ‘thinfilm phase’[19,20] with a slightly tilted angle to the substrate plane, since the molecular length of pentacene is 16.01 Å [19]. Yet the same (001) peak of the 80 °C grown film shifts to a higher degree of ∼5.92°

and it corresponds to a decreased d₀₀₁-spacing ∼14.92 Å, indicating a larger tilted degree to the substrate plane. Fur- thermore, the much lower intensity of (001) peaks of theﬁlms grown at 70 °C and 80 °C implies lower crystallinity in the

Semicond. Sci. Technol.30(2015) 035005 Y Shaoet al

(3)

vertical direction, which is highly consistent with the decreased average grain size observed in the corresponding AFM images.

The above AFM and XRD studies reveal the evolution of the pentacene microstructures manifested in surface morphology and grain size. We propose that the different crystallinity is closely related to FRC of the corresponding devices shown infigure1. The dc measurements were carried out for OTFTs with the corresponding pentacene films deposited at different substrate temperatures. Figure 3(a) shows the typical transfer characteristics of these OTFTs operating in saturation regime (source–drain voltage VDS=−30 V). The extracted carrier mobility also shows a clear dependence on the growth temperature (figure3(b)). The carrier mobility observed for RT-deposited pentacenefilm is 0.35 cm²V⁻¹s⁻¹, while it reaches the maximum value of 0.78 cm V⁻¹s⁻¹at the growth temperature of 50 °C. Above 50 °C, the carrier mobility decreases with the further increase of the growth temperature. The threshold voltage also varies with the growth substrate temperature. The V_T value shifts towards zero with the increase of the growth temperature, when the growth temperature is below 70 °C. Based on the experimental values of carrier mobility and threshold voltage, we calculated the theoretical cutoff frequencies using the aforementioned f_c expression and present the results in figure3(c) (RT, 50 °C and 80 °C). The 50 °C grown devices show the highestfcvalue and the 80 °C grown device shows

the lowestf_cvalue. The calculated results are generally consistent with the measuredf_cresults inﬁgure 1(c), conﬁrming that the changes of carrier mobility and threshold voltages critically affectf_cbehaviors. The calculated absolutef_cvalues are slightly smaller than the measured ones. According to Liu et al’s work, the experimental carrier mobility will be smaller than the real carrier mobility due to the contact resistance effect [21]. Therefore the f_c values calculated by the aforementioned f_c expression using experimental carrier mobility are smaller than the f_c values directly measured by C–f measurements. The complete reason is not yet very clear and is being further studied.

To further understand the intrinsic mechanisms responsible for the growth temperature effects on the fc

behaviors, we also analyzed the relationship between the grain sizes of the pentacene ﬁlms and carrier mobility or threshold voltages. Figures4(a) and (b) show the changes of the carrier mobility and threshold voltages with the grain sizes. Generally, the carrier mobility increases with the grain sizes. The 50 °C grown sample has the largest grain size and also the highest carrier mobility, while the smallest grain size and lowest carrier mobility were observed for the 80 °C grown sample. The effect of grain size on the carrier mobility has been discussed in several previous works [22].

Here we will not have further discussion. Unlike carrier mobility, the threshold voltages did not exhibit a clear relationship with the grain sizes, as shown in ﬁgure 4(b).

Figure 1.(a) Cross section structure of the OTFTs and MIS in dc and ac measurements; (b)C–fcharacteristics of comb-shaped OTFTs with pentaceneﬁlms grown at RT, 50 °C, and 80 °C (the channel length is 100μm and the applied gate voltage is−40 V); (c) cutoff frequency dependence on the gate voltages.

(4)

Hence, the surface morphology alters carrier mobility and threshold voltages, which finally affect the FRC of the OTFTs. Moreover, metal/semiconductor contact resistance is also an important mechanism affecting the FRC of the OTFTs [16]. We used the transfer-length method to esti- mate the contact resistance in OTFTs (figure4(c)). TheR_c value is estimated to be 3.7 kΩcm for devices with 50 °C grown pentacene film; while the estimated R_c values are 6.2 kΩcm and 5.5 kΩcm for devices with RT and 80 °C grown pentacene films, respectively. The devices with 50 °C grownfilm have the smallestR_cvalue and the highest f_cvalue, which implies that the metal/semiconductor interface may be another important factor in determining the frequency behaviors of our present devices as well as the surface morphology [10, 15]. It should be mentioned that the contact resistance is intrinsically controlled by the charge traps in the contact region, which are suggested to be

concentrated at the grain boundaries and at the metal/

organic interface.

In addition tof_c, stability of flat-band voltage (V_fb) and frequency dispersion of the capacitance–gate voltage (C–V) characteristics are also investigated to characterize the FRC of the organic devices. The stability ofV_fb is necessary to reli- ably control the turn-on of transistors [17], while the frequency stability of the capacitance is vital for the high-speed operation of the organic devices [10,14,15]. TheV_fbvalues were extracted from theC–Vcurves measured under variable frequencies (figure5(a)). All three devices exhibit reasonably good V_fb frequency stability, and the best one is the device using 50 °C grown pentacene film. The V_fb value is deter- mined by the Fermi level of the semiconductor [23], which is affected by the carrier density of the organic semiconductors.

As the 50 °C grown ﬁlm has the largest grain size, the trap density from grain boundary will be less and fewer carriers

(a)

(b) (c)

30 20 10 0 -10 -20 -30

nmnm

2.0

1.5

1.0 20 40 60 80 5 10 15 20 25

Grain Size (µm) Intensity (a.u)

Substrate Temperature ( ºC) 2Θ (deg. )

RT 40˚C 50˚C

60˚C 70˚C 80ºC

2µm

(001) (002) (003) 80ºC 70ºC 60ºC 50ºC 40ºC

RT

Figure 2.(a) AFM images of pentaceneﬁlms grown at different substrate temperatures; (b) the impact of the growth temperature on the average grain sizes; (c) XRD spectra of pentaceneﬁlms grown at different substrate temperatures. The tube anode is Cu (λK_α1=1.5406Å).

(5)

Figure 3.(a) Measured transfer characteristics of OTFTs with pentaceneﬁlms deposited at various temperatures; (b) the impact of growth temperature on the channel carrier mobility and threshold voltages; (c) the calculated cutoff frequency values as a function of the gate voltages based on the carrier mobility and threshold voltages shown inﬁgure3(b).

Figure 4.Channel carrier mobility (a) and threshold voltages (b) as function of average grain sizes of pentaceneﬁlms; (c) W-normalized transistor total resistance versus channel length of OTFTs with different pentacene grown temperatures.

(6)

will be trapped during the high frequency operation, resulting in smallVfb values and good stability. In addition, the C–V frequency dispersion of the MIS diodes was measured at frequencies ranging from 20 kHz to 100 kHz (ﬁgures 5(b)– (d)) and all the devices exhibit relatively small dispersion.

Interestingly, nearly no frequency dispersions can be observed for devices using 80 °C grown pentaceneﬁlm. This implies that the C–V frequency dispersion is probably not closely linked to the morphology such as grain size (unlikef_c andV_fb) and other mechanisms will be further investigated in the future work [24,25].

4. Conclusions

In summary, the ac frequency response behaviors of pentacene-based organic devices were investigated from the viewpoints of cutoff frequency,V_fbstability, and capacitance frequency stability. The growth temperature of pentacene ﬁlms was demonstrated to play an important role in regulating their grain sizes or carrier mobility and metal/semiconductor contact resistance, which were proved to be critical

mechanisms to affect the FRC of the MIS diodes and OFETs.

The AFM and XRD measurement results indicate that a moderate growth temperature around 50 °C∼60 °C is favorable to form well ordered crystalline structure and large average grain size. Theﬁlm grown at 50 °C has a larger grain size, higher carrier mobility and smaller interfacial contact resistance than other ﬁlms do. The corresponding devices exhibit higher cutoff frequencies for OTFTs and better V_fb frequency stability for MIS diodes. Our work provides useful insights on pentacene microstructures and their impact on the frequency response characteristics of pentacene-based organic devices.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (Contract Nos. 61271127, 51472093, 51431006) and the program for Changjiang Scholars and Innovative Research Team in University (Grant No.

IRT1243). The Program for International Innovation Coop- eration Platform of Guangzhou (No. 2014J4500016).

Figure 5.(a) Frequency dependentVfbvalues of the devices with RT, 50 °C, and 80 °C grown pentaceneﬁlms;C–Vfrequency dispersion characteristics of the pentacene grown at room temperature (b); 50 °C (c); and 80 °C (d). All the units of measurement frequencies in (b), (c), and (d) are Hz.

(7)

References

[1] Dimitrakopoulos C D and Malenfant P R L 2002Adv. Mater.

1499

[2] Katz H E 2004Chem. Mater.164748

[3] Kitamura M and Arakawa Y 2008J. Phys.: Condens. Matter.

20184011

[4] Yang Y and Wudl F 2009Adv. Mater.211401

[5] Lenski D R, Southard A and Fuhrer M S 2009Appl. Phys. Lett.

94232103

[6] Jaiswal M and Menon R 2006Appl. Phys. Lett.88123504 [7] Hamadani B H, Richter C A, Suehle J S and Gundlach D J

2008Appl. Phys. Lett.92203303

[8] Scheinert S and Schliefke W 2003Synth. Met.139501 [9] Lindner T and Paasch G 2007J. Appl. Phys.102054514 [10] Miyadera T, Minari T, Tsukagoshi K, Ito H and Aoyagi Y

2007Appl. Phys. Lett.91013512

[11] Dunn L, Cobb B, Reddy D and Dodabalapur A 2009Appl.

Phys.A95153

[12] Rotzoll R, Mohapatra S, Olariu V, Wenz R, Grigas M, Dimmler K, Shchekin O and Dodabalapur A 2006Appl.

Phys. Lett.88123502

[13] Suzuki S, Yasutake Y and Majima Y 2008Japan J. Appl.

Phys.473167

[14] Lu X B, Minari T, Liu C, Kumatani A, Liu J M and Tsukagoshi K 2012Appl. Phys. Lett.100183308

[15] Lu X B, Minari T, Kumatani A, Liu C and Tsukagoshi K 2011 Appl. Phys. Lett.98243301

[16] Meijer E J, Mangnus A V G, Huisman B H, Hooft G W, Leeuw D M and Klapwijk T M 2004Synth. Met.14253 [17] Wang A, Kymissis I, Bulovi V and Akinwande A I 2006Appl.

Phys. Lett.89112109

[18] Scheinert S and Paasch G 2004Phys. Status SolidiA2011263 [19] Dimitrakopoulos C D, Brown A R and Pomp A 1996J. Appl.

Phys.802501

[20] Chou W Y, Mai Y S, Cheng H L, Yeh C Y, Kuo C W, Tang F C, Shu D Y, Yew T R and Wen T C 2006Org.

Electron.7445

[21] Liu C, Xu Y, Ghibaudo G, Lu X B, Minari T and Noh Y Y 2014Appl. Phys. Lett.104013301

[22] Wang S D, Miyadera T, Minari T, Aoyagi Y and Tsukagoshi K 2008Appl. Phys. Lett.93043311

[23] Schroder D K 2006Semiconductor Material and Device Characterization(New York: Wiley)

[24] Knipp D, Street R A, Völkel A and Ho J 2003J. Appl. Phys.

93347

[25] Shtein M, Mapel J, Benziger J B and Forrest S R 2002Appl.

Phys. Lett.81268