Self-patterning of high-performance thin film transistors
Kuo-Jui Chang
b, Feng-Yu Yang
a,*, Cheng-Chin Liu
b, Meei-Yu Hsu
a,
Ta-Chuan Liao
b, Huang-Chung Cheng
baMaterials and Chemical Laboratories, Industrial Technology Research Institute, 195, Sec. 4, Chung Hsing Road, Chutung, Hsinchu 310, Taiwan b
Department of Electronics Engineering and Institute of Electronics, National Chiao Tung University, Hsinchu 300, Taiwan
a r t i c l e
i n f o
Article history:
Received 23 December 2008
Received in revised form 23 March 2009 Accepted 7 April 2009
Available online 14 April 2009
PACS: 73.61 Keywords: Self-patterning SAMs Surface energy
Organic thin film transistors
a b s t r a c t
We have developed a technique for the preparation of thin film transistors (TFTs) through the self-patterning of various organic and inorganic materials via solution processing using a wide range of solvents. To obtain selectively self-patterned layers, we treated the oxide dielectric with two-phase patterned self-assembled monolayers of hexamethyldisilazane (HMDS) and octyltrichlorosilane. The conducting polymer poly(3,4-ethylenedioxythio-phene) doped with poly(styrene sulfonic acid) in water and the dielectric polymer poly(vi-nyl phenol) in propylene glycol methyl ether acetate were both selectively deposited and patterned on the HMDS regions with high-quality feature shapes. When source and drain electrodes were patterned on the bottom-gate oxide wafer, we also self-patterned organic and inorganic semiconductors around the channel (HMDS) regions. These TFT devices exhibited moderate to good electronic characteristics. This method has great potential for the economical full solution processing of large-area electronic devices. The selectivity in the patterning phenomena can be understood in terms of surface energy interactions.
Ó 2009 Elsevier B.V. All rights reserved.
1. Introduction
Organic thin film transistors (OTFTs) are attracting much attention for their potential applications in next-generation devices, such as flexible displays, RFIDs, smart cards, tronic paper, sensor arrays, and low-cost disposable elec-tronic devices [1–6]. They exhibit several advantageous features – large areas, low cost, light weight, mechanical flexible, and low-fabricating temperature – that make them preferable to inorganic-based devices; in addition, they can be prepared directly on flexible plastic substrates. During the last two decades, a number of remarkable improve-ments – in materials development[7–9], the modification of self-assembly monolayers (SAMs) at organic semicon-ductor–dielectric interfaces[10–13], and device structure engineering[14–16]– have made OTFTs competitive with silicon-based devices. For practical applications, efficient patterning is crucial to the development of electronic
devices and circuits. Notably, semiconductors that are not patterned exhibit cross-talk between adjacent devices, par-asitic resistance, and gate leakage current (IG) and drain
cur-rent (ID) offsets that are more dramatic than those of
patterned systems[17,18]. There are two major approaches that are used to deposit the active semiconductor layer: thermal evaporation and solution processing. Although thermal evaporation through a shadow mask can produce well-ordered patterned films, the throughput is slow and involves expensive vacuum systems. Low-cost solution pro-cessing, on the other hand, usually produces non-patterned films that cover the entire substrate. When patterning semi-conductor films, which are usually sensitive to oxygen, water, and solvents, common photolithography methods cannot usually be applied directly without a protective cap-ping layer present on top of the semiconductor layer prior to coating of the photoresist. Parylene-C and polyvinylalcohol (PVA)[19–21]and their organic/inorganic bilayers of SiO2,
SiNx, Al2O3, and Al[22]are the most common materials used
as protecting layers. Several patterning methods have been developed recently for use in conjunction with solution
1566-1199/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.orgel.2009.04.002
*Corresponding author. Tel.: +886 3 5912643; fax: +886 3 5827694. E-mail address:[email protected](F.-Y. Yang).
Contents lists available atScienceDirect
Organic Electronics
processing, including ink-jet printing, screen printing[23], soft lithography[24,25], laser-assisted patterning[26,27], contact printing [28], and self-organization processes
[29–33]. Among these methods, self-organization has great potential for use in the fabrication of high-through-put, low-cost electronics without degrading their OTFT performances. For example, Bao et al.[29]reported a micro-contact printing (
l
CP) method for transferring low-molecu-lar-weight siloxane oligomers from PDMS stamps to pattern organic poly(3-hexylthiophene) (P3HT) and poly(vinyl phenol) (PVP) materials via selective wetting/dewetting to fabricate regions with and without oligomers[29]; never-theless, the interface of the channel region was only the bare surface, lacking a modification layer. Many authors have determined that modification layers present between the dielectric and organic active material are necessary to improve the interfacial adhesion and the film’s crystallinity, thereby providing more-stable devices exhibiting higher carrier mobilities[10]. During the preparation of this man-uscript, two research groups published self-organization processes featuring the treatment of two modification layers on silicon oxide surfaces[32,33]. Minari et al.[32] re-ported a surface presenting phenyl-terminated SAMs as channel regions and hexamethyldisilazane (HMDS) units covering the rest of the surface. Small organic molecules that had been drop-cast from toluene solutions were selec-tively crystallized on the phenyl-presenting SAM regions. Kim et al.[33]coated a hydrophobic fluoropolymer film onto silicon dioxide and then modified the channel regions with HMDS. A soluble pentacene derivative was then self-deposited from a xylene solution onto the channel areas. Although these two initial studies[32,33]appear to have several advantages over simple mono-modification treat-ment[29–31]the authors studied the patterning of semi-conductors in nonpolar solvents only; they did not explore the use of other materials (e.g., conducting poly-mers, organic gate dielectrics) or other polar and nonpolar solvents. In addition, the origins of the self-organization processes were considered to occur mostly through differ-ences in the water contact angles of the substrates inducing the hydrophilic/hydrophobic properties; there remains much room to provide a more appropriate explanation of these phenomena.In this study, we developed a new system for the self-patterning of two types of SAMs on silicon dioxide surfaces, with HMDS forming channel regions and octyltrichlorosi-lane (OTS) covering the remaining areas. We prepared these patterns readily using traditional photolithography processes. In addition to investigating several organic and inorganic semiconductors, we also processed the conduct-ing polymers from aqueous solutions and employed organ-ic gate dielectrorgan-ic materials. We explain the self-patterning phenomena in terms of surface energy differences – rather than the previously reported simple hydrophilic/hydropho-bic water contact angle theory.
2. Experimental section
Materials: PVP (Mw= 20,000), PMF (Mw= 511), OTS,
HMDS, chlorobenzene, toluene, propylene glycol
mono-methyl ether acetate (PGMEA), CH2Cl2, hexane, and
ace-tone were purchased from Sigma–Aldrich and used without further purification. Regioregular P3HT was purchased from Aldrich and purified through Soxhlet extractions with hexane and CH2Cl2to remove
low-molec-ular-weight chains. 13,6-N-sulfinylacetamidopentacene (NSFAAP) and zinc acetate were also obtained commer-cially from Aldrich and used directly. Poly(9,9-dioctylfluo-rene-alt-bithiophene (F8T2) was purchased from American Dye Source. The PEDOT:PSS water solution (AI4083) was purchased from Bayer.
Device fabrication: We prepared the substrate using the self-patterning process presented inFig. 1. We employed a silicon wafer presenting thermally grown SiO2
(d = 100 nm) as the substrate. The source and drain (S/D) Pt/Ti (50 nm/2 nm) electrodes having channel widths and lengths of 1000 and 10
l
m, respectively, were prepared using a photolithography/Pt–Ti deposition/lift-off (ace-tone) process. The entire device surface was treated with HMDS vapor and then cured at 150 °C for 0.5 h, and then the channel regions of the device were covered with pho-toresist using conventional photolithography processing. The remaining region not covered with the photoresist was treated through O2 plasma bombardment and thentreated with OTS vapor then cured at 100 °C for 0.5 h. Fi-nally, the photoresist was removed to provide a substrate presenting HMDS in the channel regions and OTS over the remaining areas.
Electrical Measurements: All TFT devices arrays in this study, each containing 20 devices were fabricated. To en-sure accuracy of data that were collected, we meaen-sured at least 10 devices for each array and no significant (<10%) variations were observed from device to device. All I–V measurements of our OTFT devices were recorded at room temperature under ambient conditions using an Agilent 4156C semiconductor parameter analyzer. The thicknesses of the corresponding films were determined through cross-sectional scanning electron microscopy (SEM). The surface energies of the various SAM surfaces were determined through contact angle measurements using a FACE contact-angle meter (Kyowa Kaimenkagaku Co.) and distilled water and CH2I2as probe liquids.
3. Results and discussion
We employed various solutions of organic and inor-ganic materials to examine the scope of this self-patterning method; Fig. 2 presents their chemical structures. After depositing a drop of the semiconductor solution [P3HT (5 mg) in dichlorobenzene (1 mL)] onto the HMDS/OTS-presenting S/D electrode-patterned substrate and then decanting the substrate to remove the large droplet, we found that some small droplets remained adhered to the HMDS regions. After the solvent had dried, the resulting films were isolated with finely featured shapes on the HMDS region. Thus, using this approach, we fabricated a bottom gate, bottom contact configuration of the patterned P3HT-OTFT.Fig. 3a displays an optical micrograph of the with uniformly patterned P3HT-OTFT device array. A mag-nified image of the P3HT-OTFT single device (Fig. 3b)
reveals rectangular features (250
l
m 300l
m). In addi-tion, when we used the HMDS/OTS-presenting substrate lacking any S/D electrodes, we formed films of the organic conductor poly(3,4-ethylenedioxythiophene) doped with poly(styrene sulfonic acid) (PEDOT:PSS) from an aqueous solution (Fig. 3c) and of poly(vinyl phenol) (PVP), a com-monly used dielectric layer, from a propylene glycol methyl ether acetate (PGMEA) solution (Fig. 3d); these polymers were also well defined in the desired regions, with their films were exhibiting uniformly striped rectan-gular featured having areas of 100l
m 1000l
m. Thus, a series of materials could be self-patterned using HMDS/OTS substrates through solution processing with a wide range of solvents.
We used an energy dispersive spectrometer (EDS) to analyze the surfaces of the patterned P3HT-OTFT devices to determine whether any residues or contaminants ex-isted outside the active region after performing the solu-tion deposisolu-tion process. Before taking SEM photographs, we deposited a thin Pt film to make the substrate conduc-tive. So the Pt signals were observed at both regions.Fig. 4
presents EDS spectra recorded over an active P3HT region and over a P3HT-free region. P3HT has the chemical for-mula C10H14S; we found S and C atom signal distributions
Fig. 1. Schematic procedure used for the self-patterning of OTFTs.
in the EDS profile of active region, as well as O, Si, and Pt atom signals representing the SiO2substrate and the S/D
Pt electrodes. In contrast, the EDS profile recorded away
from the HMDS region, i.e., the region presenting OTS, reveals signals only for O, Si, and Pt atoms, suggesting that no or non-sufficient P3HT residues or other organic contaminants were present in this area.
In addition to rr-P3HT, we also employed two other organic semiconductors – poly(9,9-dioctylfluorene-alt-bithiophene (F8T2; 20 mg in 1 mL of CHCl3),
13,6-N-sulfinylacetamidopentacene (NSFAAP; 15 mg in 1 mL of CHCl3) – and one inorganic semiconductor (ZnO precursor
solution; zinc acetate-to-ethanolamine molar ratio, 1:1; mixture concentration, 0.375 M in 2-methoxyethanol) to prepare OTFTs using the HMDS/OTS substrates. Again, we found that all of the resulting films exhibited isolated, fine feature shapes located in the HMDS regions, thereby pro-viding bottom gate, bottom contact configurations for their patterned OTFTs. Moreover, we prepared corresponding controlled non-patterned P3HT-OTFT devices to compare their off currents with those of the patterned devices.
Fig. 5 presents the electronic characteristics of the OTFT devices; we determined the transfer characteristics of P3HT, F8T2, and pentacene devices by operating the de-vices at a value of Vdsof 40 V and values of VGranging
from +40 to 40 V and of ZnO device by operating at a va-lue of Vdsof 20 V and values of VG ranging from 10 to
60 V. We define the carrier mobility (
l
) and the threshold voltage Vthusing Eq.(1):Fig. 3. Optical micrographs of silicon substrates presenting (a) a P3HT-OTFT array, (b) a single P3HT-OTFT device, (c) a PEDOT:PSS array, and (d) a PVP self-patterned array.
Fig. 4. SEM images and EDS spectra of a P3HT-OTFT device prepared using the self-patterning method.
ffiffiffiffiffiffiffiffiffiffiffi ID;SAT p ¼ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi WlCox 2L r ðVG VthÞ ð1Þ
where ID,satis the saturated drain current, Coxis the gate
capacitance per unit area, W and L are the conducting channel width and length, respectively, and VGis the
ap-plied gate voltage. We extracted the mobilities from the slopes of the linear plots of the square root of the drain current versus the gate voltage. The transfer curve of the non-patterned P3HT-OTFT device (Fig. 5a) exhibited an off current of 109A. The on/off ratio, the mobility, and
the threshold voltage were 8.9 104, 3.2 102
cm2V1s1, and 8 V, respectively. For the patterned
P3HT-OTFT device (also presented inFig. 5a), the on/off ra-tio improved dramatically from 8.9 104 to 3.8 107,
with the off current decreasing from 109to 1012A. We
suspect that the off current was suppressed effectively as a result of self-patterning. The mobility of the patterned P3HT device decreased only slightly (to 2.6 102
cm2V1s1) relative to that of the non-patterned device
(3.2 102cm2
V1s1). The differences of mobilities
caused by patterning of P3HT layer was also found by Jia et al.[17] Fig. 5b and c present the ID–VDoutput
character-istics of the non-patterned and self-patterned P3HT-OTFTs, respectively; the insets provide the drain current (ID)
off-set, defined as the value of IDat various gate biases, when
the drain voltage was zero[18]. Ideally, the value of the ID
offset would be zero when VD is zero. For the
non-pat-terned device (Fig. 5b inset), the value of the IDoffset
in-creased when higher gate voltages were applied; the ID–
VDoutput curve in the linear region was clearly distorted.
On the other hand, the ID–VDoutput curve of the patterned
device exhibited an ideal IDoffset (Fig. 5c inset), thanks to
the patterning effect of the semiconductors. Note that
NSFAAP is transformed into pentacene through thermally degradation, and that ZnO films are formed from zinc ace-tate during high temperature annealing.Fig. 5d–f display the transfer curves of the F8T2, pentacene, and ZnO TFTs, respectively. The on/off ratio, mobility, and threshold volt-age of the F8T2 device were 2 105, 2.1 105cm2
V1s1, and 5 V, respectively; for the
pentacene-contain-ing device, these values were 7 106, 3.5 102cm2
V1s1, and 7 V, respectively; for the ZnO-based device,
they were 3 103, 0.31 cm2V1s1, and 5 V, respectively.
Thus, all of these OTFT devices exhibited moderate to good electrical characteristics after performing self-patterning, suggesting that this method has great potential for applica-tion to the preparaapplica-tion of large-area, low-cost organic elec-tronic devices.
To determine the origins of the self-patterning effect, we determined the surface energies of the SAM-treated dielectrics by measuring their contact angles for distilled water and CH2I2as probe liquids and employing the
geo-metric mean Eq.(2) [34]:
ð1 þ cos hÞcpl¼ 2ðcds
c
dplÞ 1=2 þ 2ðcp sc
p plÞ 1=2 ð2Þwhere
c
sandc
plare the surface energies of the sampleand probe liquid, respectively, and the superscripts d and p represent the dispersion and polar components of the sur-face energy, respectively.Table 1summarizes the contact angles and surface energies of our various SAM-modified surfaces. The surface energies of the active regions (HMDS on SiO2: 37.4 mJ m2; HMDS on Pt: 40.7 mJ m2) of the
P3HT-OTFT devices were higher than those of the non-ac-tive regions (OTS on SiO2: 24.5 mJ m2; OTS on Pt:
31.9 mJ m2). Because the surface energy is related to the
absorbance of the surface, a surface having a high surface energy more readily absorbs the contacting liquid.Fig. 6
Fig. 5. (a) Transfer curves and (b and c) output curves of P3HT devices prepared on non-patterned and self-patterned semiconductor films. (d–f) Transfer curves of the (d) F8T2, (e) pentacene precursor, and (f) ZnO TFT devices.
presents the surface energies of the HMDS/OTS-treated surfaces with S/D Pt electrodes (stripe pattern; W = 500
l
m, L = 10l
m) as well as their relative locations. We observe that the contacting liquids more readily adhere to the HMDS surface with its higher surface energy than to the OTS surface with its lower surface energy. When we placed a drop of P3HT in dichlorobenzene onto the de-vice substrate’s surface, the droplet selectively wetted/ad-hered to the HMDS regions and dewetted the OTS regions. After decanting the substrate, the remaining P3HT solution was located (adhered) only on the areas of higher surface energy. As the solvent was evaporated, the P3HT film formed in the desired areas with confined features. The use of contacting liquids of other organic and inorganic semiconductors in either CHCl3 or methoxyethanol, ofPEDOT:PSS in aqueous solutions, and of PVP in PGMEA solution was also compatible with this method.
We note that both the HMDS- and OTS-SiO2surfaces
possessed large water contact angles, i.e., the surfaces are
hydrophobic; therefore, aqueous solutions should not ad-here to them at all. Indeed, several researchers have re-ported self-organization processes in which HMDS-SiO2 [32] and OTS-SiO2 [30,31] act as dewetting regions. In
our case, however, the two HMDS- and OTS-SiO2surfaces
provided a unique patterning platform for successful solu-tion processing using a wide range of solvents. Therefore, the simple hydrophilic/hydrophobic interactions cannot explain our wetting phenomena. We believe that the sur-face energy interactions – considering the differences in surface energy of the two SAMs – is more suitable for explaining this self-patterning phenomenon.
4. Conclusions
In summary, we have developed a solution-processable self-patterning method using a two-phase SAM-modified silicon dioxide surface for the deposition of a variety of
Table 1
Contact angles and surface energies of various SAM-modified surfaces.
Surface treatment Contact angle Surface energy (dispersion component)cd s(mJ m 2 ) Surface energy (polar component)cp s(mJ m2) Surface energy (substrate)cs(mJ m 2 ) Water(°) CH2I2(°) Bare SiO2 61.5 53.9 32.1 14.6 46.7 Bare SiO2+ HMDS 74.8 59.4 28.9 8.5 37.4
Bare SiO2+ O2plasma 25.2 43.1 38.0 32.0 70.0
Bare SiO2+ O2plasma + OTS 98.2 69.6 23.1 1.5 24.5
Bare Pt 86.9 40.8 37.2 1.6 40.8
Bare Pt + HMDS 85.2 41.9 38.6 2.1 40.7
Bare Pt + O2plasma 87.6 35.8 41.7 1.2 42.8
Bare Pt + O2plasma + OTS 95.0 56.0 30.9 1.0 31.9
Fig. 6. Surface energy diagram for the OTFT devices. 820 K.-J. Chang et al. / Organic Electronics 10 (2009) 815–821
organic and inorganic materials from both organic and aqueous solutions. The resulting films were self-patterned in the desired regions with well-defined feature shapes. From analyses of substrates presenting S/D electrodes, all of our self-patterned TFT devices exhibited moderate to good electronic characteristics. This method has great po-tential for application to the fabrication of large-area, low-cost, fully-solution-processed electronics when com-bined with mature photolithographic technology for mass production. Surface energy interaction explains this self-patterning phenomenon more suitably than does a simple hydrophilic/hydrophobic interaction.
Acknowledgments
The authors thank the Ministry of Economic Affairs (8351A11410) and the National Science Council (NSC97-2218-E-009-005) of the Republic of China for financially supporting this research.
References
[1] H. Sirringhaus, T. Kawase, R.H. Friend, T. Shimoda, M. Inbasekaran, W. Wu, E.P. Woo, Science 290 (2000) 2123.
[2] L.-L. Chua, J. Zaumseil, J.-F. Chang, E.C.W. Ou, P.K.H. Ho, H. Sirringhaus, R.H. Friend, Nature 434 (2005) 194.
[3] B. Crone, A. Dodabalapur, Y.-Y. Lin, R.W. Filas, Z. Bao, A. LaDuca, R. Sarpeshkar, H.E. Katz, W. Li, Nature 403 (2000) 521.
[4] J.H. Burroughes, D.D.C. Bradley, A.R. Brown, R.N. Marks, K. Mackay, R.H. Friend, P.L. Burns, A.B. Holmes, Nature 347 (1990) 539. [5] A. Afzali, C.D. Dimitrakopoulos, T.O. Graham, Adv. Mater. 15 (2003)
2066.
[6] A. Afzali, C.D. Dimitrakopoulos, T.L. Breen, J. Am. Chem. Soc. 124 (2002) 8812.
[7] C.D. Dimitrakopoulos, P.R.L. Malenfant, Adv. Mater. 14 (2002) 99. [8] G. Horowitz, Adv. Mater. 10 (1998) 365.
[9] A. Facchetti, Mater. Today 10 (3) (2007) 28.
[10] S. Kobayashi, T. Nishikawa, T. Takenobu, S. Mori, T. Shimoda, T. Mitani, H. Shimotani, N. Yoshimoto, S. Ogswa, Y. Iwasa, Nat. Mater. 3 (2004) 317.
[11] W.-Y. Chou, H.-L. Cheng, Adv. Func. Mater. 14 (2004) 811. [12] S.-Z. Weng, W.-S. Hu, C.-H. Kuo, Y.-T. Tao, L.-J. Fan, Y.-W. Yang, Appl.
Phys. Lett. 89 (2006) 172103.
[13] Y. Wu, P. Liu, B.S. Ong, T. Srikumar, N. Zhao, G. Botton, S. Zhu, Appl. Phys. Lett. 86 (2005) 142102.
[14] S. Natalie, R.H. Friend, H. Sirringhaus, Science 299 (2003) 1881.
[15] R. Parashkov, E. Becker, S. Hartmann, G. Ginev, D. Schneider, H. Krautwald, T. Dobbertin, D. Metzdorf, F. Brunetti, C. Schildknecht, A. Kammoun, M. Brandes, T. Riedl, H. Johannes, W. Kowalsky, Appl. Phys. Lett. 82 (2003) 4579.
[16] Y.H. Kim, S.M. Han, W. Lee, M.K. Han, Y.U. Lee, J.I. Han, Appl. Phys. Lett. 91 (2007) 042113.
[17] H. Jia, E.K. Gross, R.M. Wallace, B.E. Gnade, Org. Electron. 8 (2007) 44.
[18] H. Jia, G.K. Pant, E.K. Gross, R.M. Wallace, B.E. Gnade, Org. Electron. 7 (2006) 16.
[19] S.H. Han, J.H. Kim, J. Jang, S.M. Cho, M.H. Oh, S.H. Lee, D.J. Choo, Appl. Phys. Lett. 88 (2006) 73519.
[20] L. Zhou, S. Park, B. Bai, J. Sun, S. Wu, T.N. Jackson, S. Nelson, D. Freeman, Y. Hong, IEEE Dev. Lett. 26 (2005) 640.
[21] Y. Wang, H. Cheng, Y. Wang, T. Hu, J. Ho, C. Lee, T. Lei, C. Yeh, Thin Solid Films 491 (2005) 305.
[22] J.S. Lewis, M.S. Weaver, IEEE J. Selected Topics Quantum Electron. 10 (2004) 45.
[23] D.A. Pardo, G.E. Jabbour, N. Peyghambarian, Adv. Mater. 12 (2000) 1249.
[24] Z. Bao, J.A. Rogers, H.E. Katz, J. Mater. Chem. 9 (1999) 1895. [25] M.L. Chabinyc, A. Salleo, Y. Wu, P. Liu, B.S. Ong, M. Heeney, I.
McCulloch, J. Am. Chem. Soc. 126 (2004) 13928.
[26] D.G. Lidzey, M. Voigt, C. Giebeler, A. Buckley, J. Wright, K. Bohlen, J. Fieret, R. Allott, Org. Electron. 6 (2005) 221.
[27] H. Kobayashi, T. Shimoda, H. Kiguchi, World Pat. WO 99/46961, 1999.
[28] C. Kim, P.E. Burrows, S.R. Forrest, Science 288 (2000) 831. [29] A.L. Briseno, M. Robert, M.M. Ling, H. Moon, E.J. Nemanick, Z. Bao, J.
Am. Chem. Soc. 123 (2006) 3880.
[30] H.Y. Choi, S.H. Kim, J. Jang, Adv. Mater. 16 (2004) 732. [31] A. Salleo, A.C. Arias, Adv. Mater. 19 (2007) 3540.
[32] T. Minari, M. Kano, T. Miyadera, S.D. Wang, Y. Aoyagi, M. Seto, T. Nemoto, S. Isoda, K. Tsukagoshi, Appl. Phys. Lett. 92 (2008) 173301.
[33] S.H. Kim, D. Choi, D.S. Chung, C. Yang, J. Jang, C.E. Park, S.H.K. Park, Appl. Phys. Lett. 93 (2008) 113306.
[34] F.M. Fowkes, in: R.L. Patrick (Ed.), Treatise on Adhesion and Adhesives, vol. 1, Marcel Dekker, New York, 1967, p. 352.