Introduction

Organic semiconducting materials have recently attracted considerable attention due to their reduced cost, flexibility, wide applications, and low environmental impact. Over the last few years, performance of organic light emitting diodes (OLEDs) and organic photovoltaic cells (OPVCs) have been greatly improved by designing organic molecules which have appropriate electronic properties. With regard to organic thin film transistors (OTFTs), however, device performance is still far from practical application. What seems to be lacking is crystallinity of thin films rather than their electronic properties. The subject of this study is to control the morphology of organic thin films by applying appropriate growth condition or novel techniques. We studied organic film growth on electrode metals and insulating dielectrics, via both physical and chemical modification approaches. The target materials are α-sexithiophene (6T) and pentacene, which are depicted in Fig.1.

1: Physical modification of dielectric substrate: Graphoepitaxy of a-sexithiophene on periodic grooves

We examined epitaxy on artificially patterned substrates to fabricate high orientated organic thin film. This epitaxy is known as graphoepitaxy and is a promising technique for oriented growth on substrates with an amorphous dielectric top layer. Periodic grooves were fabricated by electron lithography on Si wafers with an oxide layer of 300nm (Fig.2). The width (W) and depth (d) of the grooves were set at 100nm and 10nm, respectively. The periodicity L has various lengths (from 200nm to 10 μm).

A typical AFM image of 6T films with a thickness of 15nm is shown in Fig.3. Three different types of 6T grains are observed; facetted grain (FG), elongated facetted grain (EG) and groove filling grain (GFGs). It should be noted that most of GFGs have EGs and FGs on them.

The crystallographic orientation of EGs and FGs can be estimated by their shapes. In-plane orientation of 6T grains can be discussed in terms of 8, which it defined as the angle between b-axis of the 6T grains and the groove direction. The histogram of in-plane orientation of EGs and FGs is depicted in Fig. 4. We can clearly see the following two important features; first, there are several peaks in histogram, indicating existence of preferential crystallographic direction. It is likely that these directions are determined to make a low index direction of the grains parallel to grooves (the assignment is shown in Fig.4). We can say that b-direction of 6T films generally tends to be parallel to the grooves for EGs. Second, 8 of EGs tends to be small (0 to 45), while that of FGs tends to be large (45 to 90). EGs would be formed as a result of FGs' coalescence on GFGs with small O. It is worth noting that connection of these EGs is perfect at the coalescence boundary, judging from the extreme flatness of EG surface.

Next, 6T morphology dependence on groove periodicity is examined. At first, densities (n) of three kinds of gains (EGs, FGs and GFGs) are evaluated. The grain density was obtained by calculating the number of grains per unit length normal to the groove, as a function of groove periodicity L. It can be seen in Fig. 5(a) that nGFG decreases gradually obeying a power law in the region of L smaller than 2 um. It is surprising that nEGandEG remains almost constant, while the site of nucleation decreased with L. This will be attributed to the suppression of nucleation around EGs and FGs. The density of EGs and FGs are determined by the capture length of 6T molecule (Capture zone model). In this model deposited 6T molecules will be immediately captured by the existing island, which results in decrease of the density of 6T migrating molecules and suppression of nucleation around the island. Therefore, nucleation-free regions appeared and the film coverage became small. The capture length Lc can be estimated to be 1 um. With this model, we can explain why groove periodicity L=2 um gives largest EG with this mechanism (see Fig.5(b)). If L is smaller than 2Lc, GFGs would be located at random positions and their capture zones are overlapped. As a result, each domain could not incorporate sufficient 6T molecules and the elongation is suppressed. When L is larger than 2Lc, on the other hand, the grain nucleation on terrace occurs and graphoepitaxy is not likely to occur. The optimized L for EGs growth is L~2Lc,≒ 2 μm.

2: Chemical modification of electrodes: growth of pentacene SAM-modified Au electrodes fabricated on SiO2 substrate

In bottom contact OTFTs, organic semiconductors are deposited on electrodes and dielectric. When the film growth on electrodes is very different from that of on dielectric, the film morphology on dielectric near the electrodes is likely to be influenced by the electrodes. The simplest approach to investigate the electric properties in these regions is to fabricate additional electric probes on the substrates, but this is not applicable since additional electrodes will also influence the film morphology. So we attempted to investigate electric properties without using any additional electrodes by use of frequency response measurement and theoretical analysis we have developed.

Two kinds of SAMs, octanethiol (OT : thiol with alkyl chain) and benzenethiol (BT: thiol with aromatic ring), were used in this study. After the fabrication of Au electrodes on Si02/Si substrates, the substrates were exposed in SAM vapor for 1 day, and then pentacene was deposited in a vacuum chamber. Device performance of fabricated samples is shown in Fig. 6. SAM treated samples showed increase in mobility and decrease in threshold voltage. Contact resistances at electrode-semiconductor interfaces are extracted by analyzing the result of frequency response measurement. Contact resistance dramatically decreases for SAM treated samples. Moreover, frequency response measurement also indicates that there are high resistivity and high off current regions near the electrodes. The spatial distribution of off-current carrier is extracted as shown in Fig. 7. Introduction of SAMs could reduce these unexpected regions. Surprisingly, these high-resistivity regions disappear completely in BT-modified samples despite the SAMs effect (on mobility, threshold voltage and contact resistance) seems to be smaller than OT-modified samples. AFM images (Fig.7 right) showed the presence of morphological disorder in a bare sample but not in a BT-modified sample. In these regions electronic states and carrier density are different from those on dielectrics. We have shown that SAMs has the effect to prevent these morphological disorders.

3: A novel dielectric substrate for organic film growth: nentacene growth on granhene oxide

Recently, graphene has attracted much attention due to its unique properties. Graphene oxides (GO), which are obtained as an intermediate substance during chemical fabrication process, are also expected as a novel dielectric of ultrathin thickness. We studied growth of pentacene film on GO substrates for aiming at device application. GO was prepared from natural graphite powder by the Hummer's method. GO flakes were dispersed in methanol. In order to obtain a single or few layers GO film, a highly doped Si substrate was dipped in the dispersion liquid and lifted. This process yielded GO sheets with various thicknesses on the SiO2/Si substrate. Topographic and current images of GO were measured with SPM measurement system (JEOL JSPM-5200).

First the I-V characteristics of GO sheets are measured using conducting cantilever of AFM measurement system. The results are shown in Fig.8. For any GO thickness, electronic conductance is well described by exponential function of F12, where F is electric field applied normal to the GO sheet. Judging from this kind of field dependence, the conduction mechanism is considered to be Poole-Frenkel emission.

Pentacene film deposited onto GO sheets consists of 3D elongated islands and 2D pyramidal layer as shown in Fig. 9(a). From the micro Raman spectroscopy measurement, it is found that orientation of pentacene molecules are lying in elongated islands while pentacene molecules are standing in pyramidal layers. The RHEED images were taken during deposition (Fig.9 (b)). The diffraction pattern is invariant under the in-plane rotation of the sample. Assuming the occurrence of transmission diffraction, the diffraction pattern was simulated. By comparing with experimental data, it is found that pentacene film orientation is likely to have preference of the (021) or (-121) planes parallel to the GO surface. These results showed the crystal structure of 3D elongated islands.

Summary

A series of studies about organic film growth on variously modified substrates were conducted in order to control organic molecular orientation. These results solved a part of complexity in OTFTs originating from fragile nature of organic materials. It should be noted that the present results are applicable to other organic molecules and provide fundamental knowledge on the development of organic electronics for practical applications.

Fig. 1 molecular organic semi-conductors studied in this work.

Fig.2 Schematic illustration of a groove pattened substrate. W and d are 100nm and lOnm, respectively.

Fig. 3 6T films on periodic groove patterned substrate (L= 200nm). (Left) An AFM topographic image. (Right) The illustration of classified grains.

Fig. 4 Histogram showing the distribution of the b-axis direction of 6T grains.

Fig.5 (a) Grain density dependence on groove periodicity. (b) Average length of EGs at variousove groove periodicity.

Fig.6 Device performance of SAMs treated pentacene TFTs. Octancethiol and benzenethiol are also depicted.

Fig. 7 (left) The density of off-current carriers. (right) AFM images of organic film disorder near Au electrodes.

Fig. 8 Poole Frenkel plot for GO sheets. The inset shows relative permittivity calculated by (eq.1).

Fig. 9 (left) The AFM image of pentacene of thickness 10nm deposited on GO sheets. (right) The RHEED image.

本論文は6章からなる.

第1章ではイントロダクションであり,有機デバイスにおける有機薄膜成長制御の意義が述べられた後,有機薄膜の面内結晶方位の制御に関する近年の研究および基板表面の自己組織化単分子膜による表面修飾とその上の有機薄膜成長の研究についてレビューを行っている.

第2章では基板表面の物理的構造を利用した有機薄膜の結晶方位の制御(グラフォエピタキシー)について述べられている.電子線リソグラフィーでSiO2/Si上に作製した溝基板上に分子性有機半導体であるα-sexithiophene (6T)を蒸着し,その薄膜成長過程を観察している.人工的に作製した溝上での伸長グレインの成長を初めて観測し,その結晶方位解析から,溝基板上における核生成・成長機構について考察している.伸長成長したグレインの生成に最適な溝周期を求め,薄膜成長がCapture Zone モデルで説明できることを示している.物理的な溝パターン上の有機半導体の伸長グレインの作製およびその成長機構から成長制御まで一貫した議論がなされており,グラフォエピタキシーを有機薄膜の面内方位制御の手法として応用するために重要な先進的な結果が得られている.

第3章では,基板表面の自己組織化単分子膜(SAM)修飾による有機薄膜の結晶性と電界効果トランジスタの特性に及ぼす効果について述べられている.チオールによるSAM処理は金電極上の分子配向を変化させるのみならず,電極近傍領域 (transition region)の薄膜の結晶性を向上させグレインサイズの拡大が起きることを明らかにした.独自に開発した周波数応答解析法を用いてSAM処理基板を用いて作製した電界効果トランジスタ(FET)の接触抵抗の減少および transition regionにおける off-current キャリアの減少を明らかにした.SAMによる有機デバイスの特性向上にはtransition regionの存在が多大な影響を与えることを見出した独自性に富む結果である.

第4章ではグラフェンの化学的合成法の中間物質である酸化グラフェンの絶縁特性を,原子間力顕微鏡のプローブ電流を用いて評価している.原子間力顕微像と電流像の同時測定から,電流像が局所的な面間伝導を反映することを明らかにしている.また,詳細な電流・電圧特性の測定から,伝導の電流-電圧特性はPoole-Frenkel 依存性を示すことを見出し,その特性から酸化グラフェンの誘電率を求めた.現在までに酸化グラフェン自体の構造は多数の報告例があるが,その伝導特性・誘電体特性などの物性は解明されておらず,本結果はその基礎的な知見を与える重要な成果と言える.

第5章では還元されたグラフェン上のペンタセン薄膜成長について議論している.基板グラフェンの還元程度により形成されるペンタセン薄膜の形態が異なることを原子間力顕微鏡観察から明らかにした.さらに,その分子配向を反射高速電子回折 (RHEED) から決定した.ペンタセン分子の配向からペンタセン-グラフェン界面構造を考察し,これらが強い π-π相互作用に由来することを指摘した.これらの結果は,炭素系材料間のエピタキシャル成長の例としての知見を与えると同時に,化学的に合成したグラフェンの有機薄膜デバイスへの電極応用の観点からも,重要な成果である.

第6章では第2-5章の総括を行い,主要な結果を要約している.

なお,本論文は池田進氏(第2章),和田恭雄氏(第2章),小幡誠司氏(第4-5章),齊木幸一朗氏(第2-5章)との共同研究であるが,論文提出者が主体となって実験,解析,考察を行ったものであり,論文提出者の寄与が十分であると判断する.

以上のように,本論文では有機薄膜の成長制御として基板表面を物理的・化学的に修飾による手法(グラフォエピタキシー・SAM)および新たな基板材料(グラフェン)を利用する方法が検討されており,それぞれについて有機薄膜成長過程を詳細に解析している.これらの結果は,今後更なる発展が見込まれる有機デバイスの応用研究において基礎的な知見を与える.また,その新規な薄膜成長機構の発見やその成長機構の探索は,結晶成長の観点からも学術的価値に富むものである.これら研究成果のオリジナリティを審査委員会一同で高く評価した.

したがって,博士(理学)の学位を授けるのに十分な資格を有すると認める.