Influence of molecular structure and microstructure on device performance of polycrystalline pentacene thin-film transistors

(1)

Influence of molecular structure and microstructure on device performance of polycrystalline pentacene thin-film transistors

Horng-Long Cheng,^a^兲Yu-Shen Mai, Wei-Yang Chou, and Li-Ren Chang

Institute of Electro-Optical Science and Engineering, Center for Micro/Nano Technology Research, National Cheng Kung University, Tainan 701, Taiwan

共Received 18 January 2007; accepted 2 April 2007; published online 27 April 2007兲

The authors have fabricated the pentacene thin films on polymethylmethacrylate共PMMA兲 and on silicon dioxide dielectric surfaces featuring similar surface energy and surface roughness. On both surfaces the pentacene films displayed high crystal quality from x-ray diffraction scans, although the film on PMMA had significantly smaller grain size. The pentacene transistors with PMMA exhibited excellent electrical characteristics, including high mobility of above 1.1 cm²/ V s, on/off ratio above 10⁶, and sharp subthreshold slope below 1 V / decade. The analysis of molecular microstructure of the pentacene films provided a reasonable explanation for the high performance using resonance micro-Raman spectroscopy. © 2007 American Institute of Physics.关DOI:10.1063/1.2734370兴 The performance of organic thin-film transistors 共OT-

FTs兲 has greatly improved lately. In particular, we can see performance improvements in conjugated oligoacenes, i.e., pentacene.^1–3 Pentacene, deposited on dielectric surfaces with surface-induced structure, forms a polycrystalline thin film with grain morphology.¹ General assumptions suggest that mobility properties appear to be dominated by grain boundary effects that occur due to the large amount of charged trapping states at the boundary.^1,2,4Most researchers utilize crystal structure and grain size to interpret the performance of OTFTs. However, some reported data are not con- sistent with the observations presented above.^3,5,6Whether in the intragrain or in grain boundary of organic film, the struc- tures still have weak van der Waals forces, unlike the inorganic polycrystalline media. However, little is known about the importance of molecular structure and microstructure within a polycrystalline organic film on OTFT performance.

Surface properties of a dielectric layer, i.e., surface energy 共␥s兲 and surface roughness 共Rrms兲, are the distinctive factors that determine potential improvements in electric characteristics of OTFT.⁷ Polymeric insulators have been considered as a preferable surface modification layer and/or gate dielectric materials due to their numerous advantages over inorganic materials. Therefore, in addition to studying the␥sand R_rms, it is also interesting to explore the influence of polar groups of polymer insulators on the structure of organic semiconductors and the corresponding performance of OTFTs.

In this letter, we have investigated the crystal structure, thin-film morphology, and molecular structure and microstructure of pentacene films grown on the polymeric and inorganic dielectric surfaces with similar␥sand R_rmsvalues.

Hence, the impact of different␥s and R_rms values could be distinguished. The study includes silicon dioxide 共SiO2, ␥s

= 49.8 mJ/ m², R_rms= 4.8 Å兲 and polymethylmethacrylate 共PMMA兲共␥s= 49.1 mJ/ m², R_rms= 5.0 Å兲 surfaces. We should note that the pentacene film, grown on PMMA surface with a very small grain size, allows the production of OTFTs with very good performance characteristics. Here, we used x-ray diffraction 共XRD兲, atomic force microscope

共AFM兲, and resonance micro-Raman measurements to dis- cuss possible triggers of enhanced mobility in pentacene- based OTFTs featuring PMMA modifications.

Heavily doped n-type Si共111兲 wafers were used as substrates and gate electrodes. A 3000 Å SiO₂ layer, thermally grown on the substrate, was used as the gate dielectric.

PMMA modified surface upon SiO₂gate dielectric was fabricated as follows: A 600 Å PMMA 共molecular weight

= 540 000, obtained from Scientific Polymer Products, Inc.兲 layer was deposited by spin coating directly onto a SiO₂ layer using a 1 wt % PMMA solution from p-xylene. The PMMA layer was baked for 4 h at the temperature of 120 ° C. Pentacene films 共Acros Organics, purity ⬃98%兲 were produced by vacuum evaporation onto dielectric surfaces with a thickness of 700 Å. The deposition rate of ap- proximately 0.5 Å / s and pressure of 1⫻10⁻⁵ torr were applied and the substrate was maintained at room temperature.

The R_rmsvalues were measured by AFM共Digital Instrument Multimode SPM AS-12VMF兲. The␥swas analyzed from the measured contact angles共FACE contact-angle meter, Kyowa Kaimenkagaku Co.兲 and the method has been described previously.⁷The pentacene films were studied by XRD in the symmetric reflection coupled ␪-2␪ arrangement. XRD pat- terns were obtained using Cu K␣ ^radiation 共␭K␣₁

= 1.5406 Å兲 and a wide-angle graphite monochromator. Ra- man spectra, produced by lattice phonons, were obtained using a Jobin Yvon LabRam HR spectrometer. A 633 nm He–Ne laser served as the excitation light source. The laser’s power was kept below 0.5 mW to prevent thermal damage of the pentacene thin film. The spatial resolution of the beam spot was around 1␮m. The resolution measurements were obtained using a 100⫻ objective microscope lens. At least three spectra were measured for each sample and every Ra- man spectrum was taken an average of 20 times. For deposited pentacene films on the same substrates, the Raman spectra are substantially the same. Electrical characteristics of OTFTs were measured by a Keithley 4200-SCS semiconduc- tor parameter analyzer in a dark vacuum chamber.

Figure 1 shows the XRD spectra of pentacene film grown on PMMA and SiO₂ surfaces. Both films have two major diffraction peaks, which can be attributed to “thin-film phase”共00l

⬘

兲 and “triclinic bulk phase” 共00l兲, respectively.^8,9

a兲Electronic mail: [email protected]

APPLIED PHYSICS LETTERS 90, 171926共2007兲

Downloaded 09 Dec 2008 to 140.116.208.51. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp

(2)

We observed the same 2␪^of 共00l

⬘

兲 peak indicating the thin- film phase with the same interlayer spacing of 15.4 Å, for both pentacene films. The XRD spectra results, combined with paracrystal theory, allow us to estimate the mean dimen- sion of the crystallites共L00l兲 perpendicular to the plane 共00l兲 and the distance fluctuation between successive planes of the family共00l兲共gII兲. Our measurements were based on the full width at half maximum 共FWHM兲 of the diffraction peaks, thus repeating the methodology used in previous reports.^8,9 The estimated L_00l_⬘and g_II values of pentacene film grown on PMMA共values for SiO2 surface are shown in brackets兲 were equal to 351± 9 Å 共352±19 Å兲 and 1.50±0.06%.

共1.48±0.14%兲, respectively. XRD analysis shows that both pentacene films have a similar crystal quality. However, we can see in the inset of Fig.1共a兲that the pentacene film has significantly smaller grain size on the PMMA surface as compared to the film on the SiO₂ surface. Similar to the previously demonstrated results, the overall nucleation density is higher on the PMMA surface than on the SiO₂surface at early stages,¹⁰ despite the difference in the experimental conditions. The observed differences between PMMA and SiO₂ substrates were attributed to the difference in the ad- sorption energy and surface diffusion of pentacene. Here, the R_rmsand␥sof the PMMA and SiO₂surfaces were almost the same. We can therefore conclude that the polar groups on PMMA surface have a great influence on the morphology of pentacene film.

The molecular structure and microstructure of both pentacene films on PMMA and on SiO₂ surface were studied using resonance micro-Raman spectroscopy. We used red excitation line 共␭exc= 633 nm兲 with the wavelength that was located precisely at the absorption peak arising from the intermolecular interactions, so-called the Davydov splitting.^11,12We expected to see a resonance enhancement of these modes related to Davydov splitting in Raman spectra.

As shown in Fig.2, we focused on the C–H in-plane bending 共1140–1190 cm⁻¹region兲 and the C–C aromatic stretching 共 1340– 1390 cm⁻¹region兲 vibrations. These modes have been assumed to contribute to the relaxation共reorganization兲 en-

ergy most intensively;¹³therefore one can expect to see the impact of the modes on the conformational transition during the carrier transport process. Moreover, there is no detectable Raman band of PMMA film on the SiO₂ surface in the energy range of 1100– 1400 cm⁻¹. The 1158 共Ag兲 and 1178 cm⁻¹共Ag兲 modes that originate from a pentacene molecule are assigned to the motion of atoms located at the end and on both sides of the pentacene molecule, respectively.^11,14 For the C–C aromatic stretching mode, there were three bands found at 1351共Ag兲, 1370 共Ag兲, and 1379共B3g兲 cm⁻¹.¹⁴First, we have noticed that the FWHM of the 1178 cm⁻¹ band and C–C aromatic stretching modes were smaller on the PMMA surface as compared to the SiO₂ surface. The smaller FWHM implied that pentacene molecule in the thin film on PMMA was more homogeneous, thus the lower molecular relaxation energy. Second, we observed opposite shifts of the frequency for the strongest C–H in-plane bending mode and the strongest C–C stretching mode. The opposite shift trends in Raman spectra have also been observed in other ␲-conjugated molecular materials during measured temperature decreases,¹⁵indicating that the unit cell volume was reduced. Third, a splitting 共⌬v兲 be- tween 1158 and 1155 cm⁻¹was larger on the PMMA surface than on the SiO₂ based surface. Previously, the bands at around 1155 cm⁻¹have been explained by Davydov splitting supported by the fact that the intensity ratio depends on the polarization.¹¹The relative intensity I₁₁₅₅/ I₁₁₅₈was equal to 1.25± 0.10 and 1.85± 0.21 for pentacene films grown on SiO₂ and PMMA surfaces, respectively. Indeed, we have found that the 1155 cm⁻¹band is related to the intermolecu- lar coupling in the direction parallel to the a-b lattice plane¹⁴ and highly sensitive in response to 633 nm excitation lines as compared to other Raman modes. The obtained results indi- cate that there were stronger interactions between the adja- cent pentacene molecules of the film based on the PMMA surface. Additionally, we have made complementary low temperature Raman measurements for experimental pentacene films共data not shown兲. The significant upward shift of the 1370 cm⁻¹band is associated with an increase in⌬v and can be viewed as evidence supporting increased intermolecular interactions. Along with the decrease in temperature, we could observe enlarged Davydov splitting in optical absorption spectra of pentacene thin-film phase. Enlarged Davydov splitting has also been predicted by Faltermeier et al.

recently.¹²

FIG. 1. 共Color online兲 XRD spectra of pentacene films grown on 共a兲 SiO2

surface and共b兲 PMMA surface. Inset: The corresponding AFM micrographs.

FIG. 2. 共Color online兲 Raman spectra of pentacene films grown on SiO2

共trace a兲 and PMMA 共trace b兲. The dashed lines serve as guidelines. The FWHMs of selected Raman bands are also shown. The spectrometer resolution is 0.2 cm⁻¹.

171926-2 Cheng et al. Appl. Phys. Lett. 90, 171926共2007兲

(3)

Next, we turn our attention to the performance of the OTFT devices produced using pentacene films deposited on PMMA and SiO₂ surfaces, respectively. Figure3 shows the typical transfer characteristics of OTFT devices. The devices, with pentacene film deposited onto PMMA surface, exhibit better performance than the devices featuring pentacene films deposited onto the SiO₂ surface. The devices featuring PMMA layer had the modulated on/off current ratio larger than 10⁶, while the subthreshold slope 共SS兲 was equal to 0.7– 1.0 V/decade. Low value of SS indicated a very low density of deep traps at the interface and the formation of the ideal field-effect channel in pentacene OTFTs with PMMA layer. The saturated field-effect mobility共␮sat兲 reached above 1.1 cm²/ V s 共with average values falling in the range of 0.8– 1.2 cm²/ V s兲. On the opposite, the devices without PMMA displayed poor performance,␮satwith values found within the range of 0.06– 0.3 cm²/ V s. Additionally, we have also fabricated gold top-contact OTFTs, for which we observed similar␮satindicating that the contact resistance has a minor effect on pentacene film growth on PMMA when op- erating at the saturation regime. Moreover, pentacene-based OTFTs may exhibit hysteresis that alters the slope of transfer characteristics.¹⁶ Here, we have not observed the significant influence of the hysteresis effects on the measured mobilities for both devices. Similarly, Uemura et al.¹⁷ have not observed the hysteresis effect of pentacene-based OTFTs using PMMA as gate dielectric due to the PMMA having disor- dered dipole moment. De Angelis et al. have shown that PMMA is a good buffer layer between pentacene and SiO₂.¹⁸ Thus, the improved device performance was attributed to better molecular microstructure of pentacene film on PMMA surface, including larger intermolecular interactions and lower molecular relaxation energy due to great environmen- tal homogeneity. Obviously, the increased density of grain boundaries does not show a negative effect on device performance.

In conclusion, this study gives clear experimental evidence that the pentacene grown on the PMMA surface had better molecular microstructure as compared to the film

based on the SiO₂ surface. However, the results of XRD scans show that the crystal structure and crystal quality of both pentacene films are virtually the same. At the same time, the film on the PMMA surface has significantly smaller grain size. Therefore, we argued that the well-ordered crys- talline structure and/or large grain size are not enough to reflect the optimized supramolecular structure of polycrystalline organic film for efficient charge transportation. Our observations are different from previous expectations¹⁹ where higher nucleation density of pentacene on PMMA surface should influence the amount of grain boundaries, which may act as charge carrier traps and consequently degrade OTFT characteristics. On the contrary, we have fabricated high performance pentacene-based OFET with a PMMA modification layer. Consequently, we believe that the study of molecular microstructure of polycrystalline organic films is an area of immense importance and interest.

The authors acknowledge financial support from National Science Council, Taiwan, through Grant No. NSC 95-2221-E-006-430. The authors also acknowledge the National Center for High-performance Computing, Taiwan.

1C. D. Dimitrakopoulos and P. R. L. Malenfant, Adv. Mater.共Weinheim, Ger.兲 14, 99 共2002兲, and references therein.

2Y.-Y. Lin, D. J. Gundlach, S. F. Nelson, and T. N. Jackson, IEEE Trans.

Electron Devices 44, 1325共1997兲.

3T. W. Kelley, D. V. Muyres, P. F. Baude, T. P. Smith, and T. D. Jones, Mater. Res. Soc. Symp. Proc. 771, 169共2003兲; S. Lee, B. Koo, J. Shin, E.

Lee, H. Park, and H. Kima, Appl. Phys. Lett. 88, 162109共2006兲.

4G. Horowitz, Adv. Funct. Mater. 13, 53共2003兲, and references therein.

5D. Knipp, R. A. Street, A. Völkel, and J. Ho, J. Appl. Phys. 93, 347 共2003兲.

6M. Shtein, J. Mapel, J. B. Benziger, and S. R. Forrest, Appl. Phys. Lett.

81, 268共2002兲.

7W. Y. Chou, C. W. Kuo, H. L. Cheng, Y. R. Chen, F. C. Tang, F. Y. Yang, D. Y. Shu, and C. C. Liao, Appl. Phys. Lett. 89, 112126共2006兲, and references therein.

8T. Minakata, H. Imai, M. Ozaki, and K. Saco, J. Appl. Phys. 72, 5220 共1992兲.

9C. D. Dimitrakopoulos, A. R. Brown, and A. Pomp, J. Appl. Phys. 80, 2501共1992兲.

10B. Stadlober, U. Haas, H. Maresch, and A. Haase, Phys. Rev. B 74, 165302共2006兲.

11T. Jentzsch, H. J. Juepner, K. W. Brzezinka, and A. Lau, Thin Solid Films 315, 273共1998兲.

12D. Faltermeier, B. Gompf, M. Dressel, A. K. Tripathi, and J. Pflaum, Phys.

Rev. B 74, 125416共2006兲.

13O. Kwon, V. Coropceanu, N. E. Gruhn, J. C. Durivage, J. G.

Laquindanum, H. E. Katz, J. Cornil, and J.-L. Brédas, J. Chem. Phys. 120, 8186共2004兲.

14H. L. Cheng, W. Y. Chou, C. W. Kuo, F. C. Tang, and Y. W. Wang, Appl. Phys. Lett. 88, 161918共2006兲; H. L. Cheng, W. Y. Chou, C. W.

Kuo, Y. S. Mai, F. C. Tang, and S. H. Lai, Proc. SPIE 6336, 63361B 共2006兲.

15B. Toudic, P. Limelette, G. Froyer, F. Le Gac, A. Moréac, and P. Rabiller, Phys. Rev. Lett. 95, 215502共2005兲.

16G. Gu, M. G. Kane, J. E. Doty, and A. H. Firester, Appl. Phys. Lett. 87, 243512共2005兲.

17S. Uemura, A. Komukai, R. Sakaida, T. Kawai, M. Yoshida, S. Hoshino, T. Kodzasa, and T. Kamata, Synth. Met. 153, 405共2005兲.

18F. De Angelis, S. Cipolloni, L. Mariucci, and G. Fortunato, Appl. Phys.

Lett. 86, 203505共2005兲.

19S. Pratontep, F. Nüesch, L. Zuppiroli, and M. Brinkmannl, Phys. Rev. B 72, 085211共2005兲.

FIG. 3.共Color online兲 Transfer electrical characteristics of silver top-contact pentacene OTFTs with 共䊐兲 and without 共䊊兲 PMMA modification layer 共channel length=120␮m, channel width= 1920␮m兲.

171926-3 Cheng et al. Appl. Phys. Lett. 90, 171926共2007兲